Principles and Guidance for Presenting Active Traffic Management Information to Drivers (2021)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

107 APPENDIX A: Full Interview Responses & Follow-Up Survey Results Full Interview Responses California respondent – Marco Ruano, marco.ruano@dot.ca.gov 1. What ATM strategies are currently deployed or being considered for deployment in your jurisdiction? • Dynamic junction control – I-5/SR 110 Arroyo Seco Parkway – this is on a substandard facility that has queuing and other safety issues – deployed pavement lighting (discontinued use after 2-3 years) and overhead signs in 2010 to control access to shoulder on the connector ramp during the afternoon peak: 3-7pm, Monday-Friday. We saw improvement in congestion and safety, but have issues based on driver perceptions on what they can and cannot do. • Dynamic lane use control – under consideration – SR 110 Arroyo Seco Parkway – developed a project study report for corridor upstream of I-5 that has many curves and substandard geometry, e.g., on-ramps have stop signs with no acceleration ramp, no deceleration lanes. This proposal would make the outside lane an auxiliary lane only during off-peak times when traffic speeds are higher and less safe for slow merging or exiting traffic with overhead signs, and some pavement markings at on-/off-ramps. • Dynamic shoulder lanes, VSL, Queue Warning – different strategies under consideration – I-105: developed a ConOps, as well as Implementation and High-Level Design Plans. 2. For each ATM strategy, please specify how is ATM information disseminated to drivers, i.e., overhead signs (one sign for each lane), overhead signs (one sign for multiple lanes), lateral roadside signs, in-vehicle displays, smartphone applications, other. In-pavement lighting and overhead and roadside signs for I-5/SR 110, not full gantries Photos, presentation: https://www.nhi.fhwa.dot.gov/resources/docs/active_traffic_mgmt.pdf 3. For each ATM strategy, what kind of signage is used? i.e., fully dynamic sign, static sign with dynamic elements, other. Fully dynamic, not full color 4. What factors were used to prioritize funding for your ATM deployments? i.e., safety benefits, mobility benefits, other. • Dynamic junction at I-5/SR 110 – deployed as part of a safety program, which looks at data driven locations and are given the higher priority • I-105 – mobility – working now to get to the project initiation phase with risks, benefits, etc. to then program the project for funding for construction, most likely from regional funding, or as part of a conversion of HOV lanes to HOT lanes, which would entail P3 funding • SR 110 Arroyo Seco Parkway – safety-driven

108 5. What resources did your agency use (e.g., documents, guidance, peer deployments) to inform the design of your ATM deployment, including signage, display format of information, content, timing, and priority for displaying information, as applicable? • Dynamic junction control at I-5/SR 110 was an in-house idea • For I-105 – we had a consultant team (Cambridge), worked with FHWA Office of Operations, and had access to other peer deployments. This project included a second major element: a framework for Caltrans to evaluate ATM strategies in various corridors, developed a tool based on deployments around the country; then use that tool to select a corridor and come up with potential alternatives for deployment. 6. How did your agency choose the types and distribution of signs (e.g., fully dynamic over every lane at ¼-mile spacing) for your deployment? Dynamic junction deployment at I-5/SR 110 was based on an idea, and did not do a systems engineering process. 7. Has your agency made any changes to the way information is presented to drivers since your ATM deployment first began operations? (e.g., different message wording, different symbols, different spacing or sign configurations, different types of signage, distance presented upstream, different thresholds in algorithms for operations, etc.) If yes, please explain. Dynamic junction at I-5/SR 110 – pavement lighting was active for 2-3 years. It was deployed to help get compliance, when the lane was closed it was lit as a solid white stripe and the overhead sign showed an X, and when the lane was open the lighting would be off (no barrier) and an arrow would be shown on the overhead sign. There was still a high violation rate, and ongoing maintenance issues with the pavement lighting that was challenging to access. As such, the pavement lighting effort had to be abandoned, but overhead signs are still working. 8. How does your agency balance the needs of drivers versus the infrastructure, operations, and maintenance costs for ATM? This is a struggle – we strive to provide as much information to drivers as possible, but there are many constraints: funding, political, environmental. Public acceptance is key to gauge the success – and will help find this balance point – how will the public react and will they understand it? Keep the public engaged through the process. Early dialogue is also needed with the highway patrol – law enforcement buy-in is key because you need compliance by the traveling public, and when you’re doing something new that drivers are not used to, you need enforcement. Dialogue with maintenance personnel is key too. 9. Has your agency evaluated the return on investment of your ATM deployments? If so, how was this done? No – we did look at crash rates and also showed some benefits for congestion. 10. If you could do the ATM deployment again, what would you do differently and why? Specifically, would you use the same method for displaying information to drivers (e.g., the types and distribution of signs: fully dynamic over every lane at ¼-mile spacing)?

109 Dynamic junction at I-5/SR 110 – A new project is underway to help manage the movement in that lane with only digital overhead signage. The intent was/is to make it more adaptive after we worked out the kinks, but had challenges with initial design. An evaluation by PATH at UC-Berkeley looked at operations and alternate strategies that might provide additional benefit using modeling. Didn’t do a systems engineering process or consult with maintenance staff, and in hindsight, that would’ve been helpful, particularly from an operations and maintenance standpoint. 11. Has your agency surveyed or received feedback from drivers, either positive or negative, about the type of information they receive or the dissemination mechanism that is used? Please explain what drivers do and do not like about the deployment. No systematic feedback; received some emails and got anecdotal feedback from California Highway Patrol. About half of CMS are used to convey travel times, and some present alternate route options on time to get to a given destination – did have a website to get feedback. Findings showed that it was less important to drivers what the exact time was, but the relative time versus their normal time – people interpret the points differently (i.e., everyone has a different context for “downtown”) 12. What types of information and dissemination mechanism do you believe is most effective for ATM? We still struggle with how to safely convey information to smartphones without it being a distraction, perhaps audio is one-way? Improved roadside signage is being deployed, and the new digital overhead signs are effective. Still unsure of the effectiveness of pavement markings. 13. Has your agency given any consideration to in-vehicle or smartphone application messages for deploying ATM in a connected vehicle environment? Has your agency considered phasing out the deployment or operations of traditional ITS or on-road signing, such as DMS? If yes, please explain. Caltrans is still deploying new ITS infrastructure. It is still a challenge for Caltrans to maintain infrastructure; we still are using 50-year old loops in places and have not been able to deploy automated detection methods yet – we are working to get smarter. Nothing is being phased out, but as technology evolves, Caltrans deploys updated devices. Illinois respondent – Juan Pava, Work Zone Safety Engineer, juan.pava@Illinois.gov 1. What ATM strategies are currently deployed or being considered for deployment in your jurisdiction? Temporary systems for work zones: queue warning systems, travel time estimation, dynamic alternate routing with ITS based on congestion levels; have considered dynamic merge systems to get rid of concerns with zipper merge

110 Permanent systems: reversible lanes on I-90, I-90 Tollway SmartRoad under construction for dynamic lane control and queue warning opening 2017, signs over each lane; shoulder running for buses [NOTE: RESPONSES BELOW ONLY FOR WORK ZONE SYSTEMS] 2. For each ATM strategy, please specify how is ATM information disseminated to drivers, i.e., overhead signs (one sign for each lane), overhead signs (one sign for multiple lanes), lateral roadside signs, in-vehicle displays, smartphone applications, other. Generally, fully dynamic PCMS on the side of road are used. Have incorporated overhead DMS, and also used static signs with flashing beacons – “real-time traffic control system,” e.g., “Slower traffic ahead when flashing.” For a queue warning system, deploy as many PCMS as necessary to span the length of the queue, based on a queue analysis that is done to determine how far the queue will reach. Place dual PCMS on both the inside and outside shoulders, and develop a project-specific design for deployment. We require them to be synched, but it is very difficult to control the synchronization, so now stagger them by about 500’ so that they are not directly side by side. If they do get out of synch, it can be very confusing as drivers try to read both signs. Staggering is also helpful if a car is in the middle lane between large trucks, staggering raises the chance that they can read one of the PCMS. 3. For each ATM strategy, what kind of signage is used? i.e., fully dynamic sign, static sign with dynamic elements, other. (see above) 4. What factors were used to prioritize funding for your ATM deployments? i.e., safety benefits, mobility benefits, other. -Primary reason for queue warning systems is safety: to reduce rear-end crashes. By reducing crashes, there is a benefit to mobility also – if excessive queues (a mobility issue), that triggers the need for queue warning to avoid crashes (safety). -Have considered dynamic merge systems to reduce safety concerns with zipper merge. -For dynamic alternate routing and travel times – look at mobility benefits. -Districts have on-call contracts for deploying ITS in short-duration projects, typically less than 2 weeks, where the total cost of the project may not warrant ITS, but there is potential for unexpected queues. These are almost exclusively used for queue warning systems. 5. What resources did your agency use (e.g., documents, guidance, peer deployments) to inform the design of your ATM deployment, including signage, display format of information, content, timing, and priority for displaying information, as applicable? FHWA provides good guidance for the deployment of ITS systems: The Work Zone ITS Implementation Guide. Look at expected number of sensors and PCMS, but also determine the final layout between service provider and IDOT staff, and can make any needed changes before deployment. We rely heavily on the ITS vendors, because they’re very aware of what they are able to provide and what their systems are capable of achieving.

111 We typically use the system-provided recommendations for messages to post on the PCMS – if there is no congestion, post something like “left lane closed ahead, use caution”; when speeds drop below a certain threshold, post “slow traffic ahead”; when speeds are under 25 mph, post something like “slow traffic ahead, be prepared to stop.” 6. How did your agency choose the types and distribution of signs (e.g., fully dynamic over every lane at ¼-mile spacing) for your deployment? For a queue warning system, we deploy as many PCMS as necessary to span the length of the queue, based on a queue analysis that is done to determine how far the queue will reach. Also, a study with University of Illinois looked at the performance of ITS systems based on different sensor types and reliability of sensors. 7. Has your agency made any changes to the way information is presented to drivers since your ATM deployment first began operations? (e.g., different message wording, different symbols, different spacing or sign configurations, different types of signage, distance presented upstream, different thresholds in algorithms for operations, etc.) If yes, please explain. We have not changed much. However, IDOT still has not finalized a specification; we have a spec that we continue to tweak based on lessons learned from our experiences. We generally leave responsibility to the individual districts on the displayed message – if they hear feedback or complaints, their engineer can suggest a different type of messaging. There isn’t standard messaging. On I-74 near Champaign we tried to provide queue warning, travel time, and alternate routes on the same PCMS – had a very complicated logic for the PCMS, which led to providing confusing messages because of the large message set – how much information can you give to the drivers in a clear, concise way to give all the information necessary? We reverted back to a queue detection system primarily rather than do all 3. Doing all three can be done in the future for the same project, but we think next time we would need signs to be dedicated for each type of system. 8. How does your agency balance the needs of drivers versus the infrastructure, operations, and maintenance costs for ATM? This is a difficult question to answer for these ITS systems because the purpose of deploying these systems is for incident mitigation or mobility improvement, and it’s difficult to gauge something that isn’t happening – i.e., these systems help to prevent crashes. A need of the driver is to have reliable travel times, so we assume the cost of providing ITS is worth the theoretical benefit of deploying based on the improved mobility due to reduced crashes. 9. Has your agency evaluated the return on investment of your ATM deployments? If so, how was this done? Nothing formal – this is very difficult to do, especially with work zones because they are short term and conditions are different for different locations, etc. Ted Newsy in District 8 did a rough calculation on the number of crashes before and after the deployment of ITS, but we have not conducted a thorough benefit-cost analysis.

112 10. If you could do the ATM deployment again, what would you do differently and why? Specifically, would you use the same method for displaying information to drivers (e.g., the types and distribution of signs: fully dynamic over every lane at ¼-mile spacing)? 11. Has your agency surveyed or received feedback from drivers, either positive or negative, about the type of information they receive or the dissemination mechanism that is used? Please explain what drivers do and do not like about the deployment. We have not conducted any formal survey or study on user satisfaction or perception. Different districts have heard feedback informally, usually negatives: complaints from drivers who were not sure of what the message meant or timeliness of the message being relayed to the PCMS, e.g., delay on sign displayed 30 minutes, but it was actually 45 minutes, or sign says there is stopped traffic ahead when there was no issue. 12. What types of information and dissemination mechanism do you believe is most effective for ATM? On-board messaging would be great, but that is far into the future. Regarding PCMS, have heard anecdotally that people may be starting to get numb to them, so there is a push from some districts to use static signs with flashing beacons instead – drivers may be more likely to read the sign if they see flashing. ATM information goes beyond road users, it is also important for the agency to know what is going on – it is a public information tool, but also performance management infrastructure. 13. Has your agency given any consideration to in-vehicle or smartphone application messages for deploying ATM in a connected vehicle environment? Has your agency considered phasing out the deployment or operations of traditional ITS or on-road signing, such as DMS? If yes, please explain. There is a lot of talk in the industry re: DSRC and WAZE, but IDOT is not considering deployment; it is an immature technology and we don’t want to add another layer of distraction before it has been proven. We are not phasing out or slowing down deployment of traditional ITS. Minnesota respondent – Brian Kary, brian.kary@state.mn.us 1. What ATM strategies are currently deployed or being considered for deployment in your jurisdiction? Temporary • Dynamic lane reversal – now using moveable barrier for construction project on a 6-lane segment of I-94 that’s being operated as 5-lanes during construction; always keep three lanes in a peak direction; had a similar project on a 4-lane segment of I-494 being operated as three lanes. • Dynamic shoulder use – the segment opened on I-35W northbound in 2009, but is going away next year during construction to re-build the freeway with more lanes, and will not be in place afterward.

113 Permanent • Dynamic lane reversal – on I-394 since 1991, barrier-separated – two lanes for about 3 miles, operates as a MnPass HOT lane, controlled with gate arms. • Dynamic Lane Use Control on I-35W since 2009, I-94 since 2011-2012: o Variable advisory speeds used on these systems but turned off about 2 years ago, because we saw low compliance and few or no benefits related to throughput or incidents; the system was not responsive to the backups and o Queue Warning – testing this approach more now – looking to do similar applications that can be done with a single sign instead of each lane. 2. For each ATM strategy, please specify how is ATM information disseminated to drivers, i.e., overhead signs (one sign for each lane), overhead signs (one sign for multiple lanes), lateral roadside signs, in-vehicle displays, smartphone applications, other. • MnPass signs are one overhead sign, generally over MnPass lane • Currently use one overhead sign for each lane; scaling back to one overhead sign in some places. 3. For each ATM strategy, what kind of signage is used? i.e., fully dynamic sign, static sign with dynamic elements, other. • MnPass signs are static signs with dynamic elements • Lane control signs are full-color, fully dynamic 4. What factors were used to prioritize funding for your ATM deployments? i.e., safety benefits, mobility benefits, other. Safety and mobility both. I-35W was funded by UPA to deploy innovative technologies. As a matter of policy, MnDOT considers ATM strategies before expansion to improve safety and mobility – doing so is a high priority for the department. 5. What resources did your agency use (e.g., documents, guidance, peer deployments) to inform the design of your ATM deployment, including signage, display format of information, content, timing, and priority for displaying information, as applicable? We worked with WSDOT a lot because they were doing a similar ATM system at the same time and looked at the FHWA scanning tour documents a lot from Europe. We struggled because MnDOT was on the cutting edge of deploying this technology. We had a peer exchange with the UK agency who had deployed a similar system because they were interested in learning about MnPASS and flew in during MnDOT’s ATM design phase. It was a challenge to meet MUTCD compliance, and expand on those design standards with new ideas. We also worked with the regional FHWA office to come up with a messaging approach of our own. Now we would have more peer interactions to see what is working for other agencies, since more agencies have deployed VSLs, e.g. 6. How did your agency choose the types and distribution of signs (e.g., fully dynamic over every lane at ¼-mile spacing) for your deployment? (see response to question 5)

114 7. Has your agency made any changes to the way information is presented to drivers since your ATM deployment first began operations? (e.g., different message wording, different symbols, different spacing or sign configurations, different types of signage, distance presented upstream, different thresholds in algorithms for operations, etc.) If yes, please explain. • Added the word “closed” to the red X symbol. • Moved away from using the yellow X and now just display the words: “lane closed ahead.” • Initially provided symbols about an incident 1.5 miles upstream, now just use the signs immediately upstream of the incident: queue jumpers upstream during low speeds degraded the benefit of the further advanced notification. • The ATM design for I-94 moved forward before I-35W ATM was completed, so it has the same basic and operations. • Replacing signs on I-94 westbound with single sign instead of signs for each lane; the vendor went out of business and the signs are not lasting as long as expected, decided on this approach because we think it may provide a similar benefit. 8. How does your agency balance the needs of drivers versus the infrastructure, operations, and maintenance costs for ATM? Why invest in these ATM systems instead of incident response vehicles or capacity? The ATM lane control signage is a Cadillac system with a high cost to install and maintain, and so we are cutting back to fewer signs and longer spacing. For what we pay in added utilities costs, we could have had two additional trucks for each corridor instead of just the one; are we better off being able to display an X ahead of the incident scene or better to have an Incident Response Team (IRT) vehicle respond to clear the incident sooner? We know our incident response trucks have a 15:1 BCA; no BCA on ATM because we know the cost is very high, and it is hard to measure quantifiable benefits. 9. Has your agency evaluated the return on investment of your ATM deployments? If so, how was this done? A benefit-cost analysis was done for I-35W as part of the UPA Evaluation, but most of the benefit was due to the expanded capacity at the bottleneck. Have not done a targeted BCA for the ATM signs, the costs are high for deployment, operations, and maintenance, it is hard to measure quantifiable benefits, and we are not likely to see a huge benefit relative to those costs. Conducted a study about having a MnPASS lane on SR 77, and examined reversible lanes vs. adding a lane; the ongoing O&M of the moving barrier made it cheaper to add a lane in the long run. 10. If you could do the ATM deployment again, what would you do differently and why? Specifically, would you use the same method for displaying information to drivers (e.g., the types and distribution of signs: fully dynamic over every lane at ¼-mile spacing)? As we are doing now on I-94, we think we may be able to provide similar benefits with single signs instead of one over each lane. We display the same speed over all lanes anyway, so that could be put onto a single sign. Our ATM system does not display why the speed was lower, which could impact compliance and also add to driver confusion, i.e., “is that the speed downstream or the speed you want me to go now?” Instead, we could display “reduce speed to 35 mph.” For incidents, there was some benefit to have the red X over the lane, but

115 for a single sign, we are considering a displaying pictogram with a red X on a lane to show it is closed. 11. Has your agency surveyed or received feedback from drivers, either positive or negative, about the type of information they receive or the dissemination mechanism that is used? Please explain what drivers do and do not like about the deployment. • We conducted online driver surveys to get an initial driver understanding for messaging – closed message with red X, and due to some uncertainty, we added the word “closed” beneath the symbol; and understanding of chevron for merge. • Initially used a yellow X, but went away from that due to some driver uncertainty, and now just use words “lane closed ahead”. • UMN (Kathleen Harder) conducted a human factors study to get driver reactions to the merge chevron symbol in a simulator. • UMN (John Hourdos) looked at videos of traffic vacating the lane when a merge chevron was displayed, and drivers did vacate the lane. • We conducted research to develop VSLs and did an analysis on expected benefits. • One study did find some benefit after the deployment, but the benefit was at the minor bottleneck (the third largest bottleneck), and so there could be some confounding factors. • Conducted a human factors study to examine angle arrow vs. chevron for merge; have seen success with this symbol. 12. What types of information and dissemination mechanism do you believe is most effective for ATM? Timely and accurate, clear information (e.g., regarding the dynamic advisory speeds we posted, there was confusion about what the displayed 35 mph means – do you want me to go 35 mph now or is that the speed of traffic ahead?) 13. Has your agency given any consideration to in-vehicle or smartphone application messages for deploying ATM in a connected vehicle environment? Has your agency considered phasing out the deployment or operations of traditional ITS or on-road signing, such as DMS? If yes, please explain. We have started to talk about it – have started communicating from snow plows to infrastructure; the MnDOT 511 app has the ability to provide hands-free messaging as you’re driving, e.g., roadwork or congestion is on the roadway ahead of you – there are still some kinks that need to be worked out with this, but it is a first step in that direction. Is there a need to invest in large ATM deployments when there could be more in-vehicle apps that can do it more cost effectively? MnDOT is still deploying traditional ITS, although the ITS network is already robust in metro area. That said, there will not be a saturation rate for in- vehicle messaging to connected vehicles within 15 years, which is at the end of the lifecycle of any DMS deployed now; so, we deploy the DMS now, but may not replace it, depending on the state of connected vehicles at that time. Oregon respondent – Dennis Mitchell, dennis.mitchell@dksassociates.com 1. What ATM strategies are currently deployed or being considered for deployment in your jurisdiction?

116 • Dynamic lane use control – permanent; capabilities in place, not used • Queue warning – permanent • VSL – permanent and temporary deployment ODOT has both regulatory and advisory VSL systems. The capability for lane use control is not used – pre-programmed messages may be used at a future time, e.g., red X’s. VSL systems are for weather, recurring congestion, and incidents. OR 217 was going to be congestion based, but received federal funding to incorporate operations for weather; the software used by all the systems was written for both congestion and weather. The Mt Hood system will be mostly for weather but also congestion. ODOT has used portable trailer-mounted DMS for a VSL in work zones; did it as a pilot but it is a solution that is still available. 2. For each ATM strategy, please specify how is ATM information disseminated to drivers, i.e., overhead signs (one sign for each lane), overhead signs (one sign for multiple lanes), lateral roadside signs, in-vehicle displays, smartphone applications, other. • In Portland, there are separate signs over each lane for the advisory VSLs. • The regulatory VSL is on a two-lane roadway and is on the side of the road. • The Mt. Hood VSL is on a cantilever over the shoulder even though it is two-lane roadway, due to snow blowers that could block it from view. • Eastern Oregon system has overhead signage: one sign over two lanes. 3. For each ATM strategy, what kind of signage is used? i.e., fully dynamic sign, static sign with dynamic elements, other. • Fully dynamic signs are on I-84, OR 217, Mt Hood, and at locations under construction on US 26 from I-405 to OR 217 and I-84 from I-5 to I-205. • First deployments on I-5/I-405 in downtown and at Staley’s Junction used a static sign with a changeable number. 4. What factors were used to prioritize funding for your ATM deployments? i.e., safety benefits, mobility benefits, other. • Safety is the first priority – reduced crashes. • Variability in travel time is important – trying to reduce travel time is hard to do. • After the initial deployment on SR 217, were able to use the findings on crash reduction to get TIGER grant funding for other deployments on US 26 and I-84 and safety funding on I-5 based on the reduction in crashes. 5. What resources did your agency use (e.g., documents, guidance, peer deployments) to inform the design of your ATM deployment, including signage, display format of information, content, timing, and priority for displaying information, as applicable? Staley’s Junction was first – a weekend problem with people driving to coast and a new state park. It was a problematic T-intersection and an interchange was to be built but deferred; VSL was a temporary solution there. ODOT releases grant money to do innovative projects – got money for I-5/405 for static VSL on freeway.

117 No specific documents, but did look at peer deployments. Looked at Seattle and lessons learned from other states. When OR 217 had an opportunity to do a VSL system, we looked at what other states were doing – brought in other states for a peer review, had focus groups with random drivers to see what they could understand and how to explain what they meant. Whether you do advisory or regulatory – we learned it is easier to go from advisory to regulatory than to step back. Seattle tried hard to not overpromise and not underdeliver – initially saw a big drop in crashes, but later was not as good as initial – it’s important to manage expectations that these systems won’t solve everything or fix congestion. I-5/I-405 was designed to be regulatory, but it was never turned on as such; it was re-configured as advisory. Regulatory/advisory – allay fears that it’s a speed trap. How do you market it so people understand? Gauge public acceptance of system. A lot of people think it’s a great idea because of all the other drivers out there (i.e., I’m ok, but others need it). ODOT developed a video and website to help people understand the systems, did a media campaign and news stations did stories, and explained the systems, potential benefit, and help drivers know what they needed to do. Another lesson we learned: in Seattle, they display a minimum speed to show on their system. ODOT will not show a minimum speed – if it drops below 30 display “slow” instead of a speed to reduce driver confusion about a displayed speed that is higher than they can travel. We talked to law enforcement: Oregon’s basic rules say drivers can be cited for unsafe conditions – this is technically easier than issuing a citation for exceeding the speed limit because the officer does not need to know the sign display. On OR 217, we saw that people did not all go the speed advised, but since crashes dropped, you realize they do pay attention. 6. How did your agency choose the types and distribution of signs (e.g., fully dynamic over every lane at ¼-mile spacing) for your deployment? No fixed distance between signs, but looked at: where we could fit them in, the back of the queue for congestion points, and installed signs in front of those points for queue warning. We found where there would be good spacing between existing signs and these new signs AND where we could best make use of existing structures – new structures are the biggest cost, so use overpasses as much as possible to minimize costs. Spacing is a problem on OR 217 where it’s a short section with many interchanges. Chose fully dynamic signs so they could replicate any static sign – to do either regulatory or advisory, but also other options – we built in additional stock message sets to the software for future lane use control applications, if desired. 7. Has your agency made any changes to the way information is presented to drivers since your ATM deployment first began operations? (e.g., different message wording, different symbols, different spacing or sign configurations, different types of signage, distance presented upstream, different thresholds in algorithms for operations, etc.) If yes, please explain. • I-5/I-405 regulatory signage was changed to advisory before it was activated.

118 • The software included parameters that could be adjusted; some changes have been made to allow for more parameters that can be adjusted, but no major changes to the system – the same software is used for all deployments. • On I-84, used lasers to look at the pavement condition to develop a grip factor for friction between tire and pavement and slow speed limits. With two lanes in each direction, there is different usage in different lanes. For instance, if it is really snowing, most people are in right lane. The laser might see bare pavement in ruts of right lane, but not the left lane – this is a challenge for what speed to display – should we have a laser sensor in each lane? 8. How does your agency balance the needs of drivers versus the infrastructure, operations, and maintenance costs for ATM? The biggest driver is a reduction of crashes – this is the ultimate need of the driver! These systems aren’t cheap, but they are cheaper than building new lanes on the freeway. Infrastructure is already in place for communications and operations, so it’s a relatively small cost, comparatively. We cannot eliminate congestion, but we can impact the crashes. 9. Has your agency evaluated the return on investment of your ATM deployments? If so, how was this done? Yes – using crash data. We will receive funding in the next STIP to deploy VSLs on a section of I-5 from downtown to the Washington state line using safety dollars, which is based on benefit/cost analysis. Available funding is intended for smaller projects, but this ATM project was based on ROI using the dollar value of the reduction in crashes – it makes a good ROI. Did have some reduction in travel time and increased throughput during peak hour on OR 217, but didn’t go through an extensive study to see the value. Saw a 21% reduction after first 6 months of crash data; and it was still a 20% reduction after a full year. Never hinted about possible improvements in travel time when initially deployed, but did occur. 10. If you could do the ATM deployment again, what would you do differently and why? Specifically, would you use the same method for displaying information to drivers (e.g., the types and distribution of signs: fully dynamic over every lane at ¼-mile spacing)? What we’ve done has worked and is working well. There is nothing for the Portland-area deployments that would do differently. For the I-84 deployment, might put lasers in both lanes to measure bare pavement. 11. Has your agency surveyed or received feedback from drivers, either positive or negative, about the type of information they receive or the dissemination mechanism that is used? Please explain what drivers do and do not like about the deployment. ODOT didn’t do a survey, but we do get feedback – mostly positive now. ATM includes a lot of things in the state. When we initially put travel times on all DMS in the area, people wouldn’t believe the times that were on the sign: the time was what was calculated for that moment, and conditions could change as you drive to your destination. But now people cannot believe how accurate they are. Now ODOT has signs on arterials leading up to the freeway, which includes the travel time to the freeway and ramp meter time. When one of those signs was down for a week for repair, we received calls from people asking when that sign would be back up because they use that info for making decisions.

119 12. What types of information and dissemination mechanism do you believe is most effective for ATM? Full-color LED dynamic signs light up and get people’s attention versus a static sign with dynamic elements: those signs do command attention. These deployments are based on safety; given an advisory speed, the driver might not see the congestion ahead, but ODOT steps the suggested speed down slowly because you don’t want a big drop. The fact that the driver sees the message is different than normal (i.e., blank is default), lets them know to be alert and more prepared as they continue down the road – this results in reduced crashes. 13. Has your agency given any consideration to in-vehicle or smartphone application messages for deploying ATM in a connected vehicle environment? Has your agency considered phasing out the deployment or operations of traditional ITS or on-road signing, such as DMS? If yes, please explain. ODOT is anticipating for cars to provide data to ODOT systems, and to give information to drivers in the vehicle somehow. It takes time to get that deployed to every vehicle, so ODOT wouldn’t stop doing traditional deployments in the near future, but is instead planning for integration with it. ODOT is actively looking for ways to make that integration easier – follow the standard to providing information, but not looking to be the disseminator of that information in the vehicle: the equipment in the vehicle is not ODOT’s job. ODOT would provide the information in a standard format, and others would take it and provide it. ODOT is still installing new DMS and other traditional ITS. Utah respondent – Joshua Van Jura, jvanjura@utah.gov 1. What ATM strategies are currently deployed or being considered for deployment in your jurisdiction? VSLs – temporary deployments in work zones. • Background: there are a lot of 80 mph speed limits in Utah, I-215 around Salt Lake City is 70 mph. The flat, straight roadways can make 80 mph seem slow. • Objective: get a significant reduction in speed in an active work zone, predominantly for worker safety. We relate it to the concept of a school zone – a lot of people really slow down to 20 mph when the speed zone is active and people see kids. That said, you don’t comply with a flashing sign in June if you don’t see kids, and the same is true with work zones. • Design: with a gut feel, chose 1.5-2-mile lengths to reduce speeds from 80 mph to 45 mph. Have done a VSL for as short as 800 yards – and saw fantastic compliance for that. Generally, just one VSL sign is active. • Speed determination: We set the speed limit based on driver input: people are going to drive at whatever speed they feel comfortable with. The posted VSL algorithm measures the average speed within that 1.5-2-mile work zone, and then posts a VSL that is 10 mph under that average speed. The algorithm starts to stabilize, and has exclusions for low volumes. If the average speed is under 35 mph, it is defined as a queuing condition and the VSL defaults to a minimum speed. Low and high-speed thresholds are set; for high end: never post a speed above the posted speed, but more often never post above 10 below posted speed (70 mph) when workers are present; for low end: minimum speed is also set

120 (45 mph). Don’t change VSL display more often than 15 minutes; this is what the permanent, weather-based VSLs use. • May use static signs if there’s a double drop – 80 to 70 first, don’t drop the speed limit by more than 25 in a single drop. • Equipment: Initially purchased 4 VSL trailers for SMRT work zones, a contractor procured another seven for a separate project and left to UDOT, then got federal aid grant to rent seven more trailers for 2 years (cannot retain equipment with the federal grant). • Operations: generally, have static signage about the work zone, then a VSL at the beginning of the active work area, and a second VSL at end, and leap frog the first sign downstream to move with the active work zone. The signs know what order they are in based on lat/long coordinates, and both are always tied to a portable DMS that is upstream to know what direction they are facing. Some trailers have a feature to make the brake lights on the trailer blink when the radar detects vehicles that are driving 10 mph over the speed limit, but this feature is prohibited by the MUTCD. Have permanent weather VSLs also. Also use radar feedback signs – “Your speed is…” – we have never done studies to see they work, but observational data indicates that people become accustomed to them after about a week, e.g. urban arterial projects with commuters – so we move it to a different location within the project – it is a cheap technology to deploy for a potential benefit. Same for campaigns like “slow down, my daddy works here”: don’t have anything to show they actually work, but it cannot hurt. 2. For each ATM strategy, please specify how is ATM information disseminated to drivers, i.e., overhead signs (one sign for each lane), overhead signs (one sign for multiple lanes), lateral roadside signs, in-vehicle displays, smartphone applications, other. Lateral roadside sign on right side of road, for both permanent and work zones, which are on trailers 3. For each ATM strategy, what kind of signage is used? i.e., fully dynamic sign, static sign with dynamic elements, other. Static signs with dynamic numbers – for permanent and temporary deployments 4. What factors were used to prioritize funding for your ATM deployments? i.e., safety benefits, mobility benefits, other. Safety – particularly worker safety. These VSLs are a “huge win for minimal money.” 5. What resources did your agency use (e.g., documents, guidance, peer deployments) to inform the design of your ATM deployment, including signage, display format of information, content, timing, and priority for displaying information, as applicable? We reached out to variety of Utah stakeholders in interviews when developing the Concept of Operations – traffic engineers, construction personnel, and construction companies – tried to weight the wants and needs

121 6. How did your agency choose the types and distribution of signs (e.g., fully dynamic over every lane at ¼-mile spacing) for your deployment? (see responses to question 1, 5, 7) 7. Has your agency made any changes to the way information is presented to drivers since your ATM deployment first began operations? (e.g., different message wording, different symbols, different spacing or sign configurations, different types of signage, distance presented upstream, different thresholds in algorithms for operations, etc.) If yes, please explain. • A contractor thought deploying VSLs in 5- to 4-lane work zone on a big project for I-15 could help with speed compliance, but it was not effective. • At first would only change speeds at the beginning and end of working shift. • Originally only 1.5 miles long, but added flexibility to the deployment design; in general, discourage more than 2.5 miles of the reduced speed zone for VSLs. • We are currently researching whether VSL have better compliance because they do not look like a typical speed limit sign, and have better visibility with the digital display being seen by the driver from further away. Currently doing a compliance study to see if we can get similar results with VSLs from static signage. • As part of a federal grant, we are going to deploy 7 VSL trailers in one-mile increments along a project and only move the sensor trailer, instead of our typical leap frog approach. With the positioning data and communications, the signs will know to only reduce the speed where it is actually needed. We set them up in defined pattern and use cellular communications. All VSLs displays would be on as 70 mph except the one section where it drops to 45 mph. • We are interested in reducing the time between posting different speeds from 15 minutes to 5-10 minutes. What is too often that you confuse drivers? There is also an enforcement issue: how do they know? We have the ability to send a text message, but the same law enforcement aren’t there often enough to want that. As for setting the posted VSL: at what point does speed variance increase accident rates? Once you get outside of +/- 10 mph, you start to increase crash rates – don’t want to make it less safe for drivers. • Going to test different VMS messages this summer as part of a queue warning system – limited on rural 2-lane to 1-lane projects – “60 seconds for our safety” will be the general pitch when no queue message is displayed to emphasize the small delay added for worker safety: a lot of these projects are so remote that you are at least an hour away from your destination. Deploying three Doppler sensors – one just upstream of lane drop, second ½-1 mile upstream, and a portable DMS ½-1 mile upstream of that – just over a mile to just over 2 miles upstream will be the sensor array. If traffic speeds are 35-55 mph, post “slow traffic ahead”; if less than 35 mph, “prepare to stop”; the default message on the sign now is “road work ahead,” but will change that to the more catchy message to encourage slower speeds to see if get better compliance. 8. How does your agency balance the needs of drivers versus the infrastructure, operations, and maintenance costs for ATM? Would say we don’t – we are mobility based. We heavily rely on user costs and anything can be justified by user costs and delays: if we can increase mobility for taxpayers and citizens, we are all for it.

122 9. Has your agency evaluated the return on investment of your ATM deployments? If so, how was this done? No – it is hard to calculate a “reduced crash” in a work zone, because they are generally so short term. Anecdotally, we had a triple semi-truck trailer enter a work zone, workers had to scramble – it was a close call that would’ve been far worse at a higher speed – this demonstrates benefit to us. Work crews find the VSLs to be effective. 10. If you could do the ATM deployment again, what would you do differently and why? Specifically, would you use the same method for displaying information to drivers (e.g., the types and distribution of signs: fully dynamic over every lane at ¼-mile spacing)? Still exploring new possibilities to learn what works – these VSLs are part of a new approach and strategy to gain public confidence: you need to slow down and we mean it, we are not just trying to make your life miserable by making you slow down for all 8 miles of a work zone, but only where you need to because there are workers present. 11. Has your agency surveyed or received feedback from drivers, either positive or negative, about the type of information they receive or the dissemination mechanism that is used? Please explain what drivers do and do not like about the deployment. No 12. What types of information and dissemination mechanism do you believe is most effective for ATM? It’s very important to maintain public confidence: do not give false information, give complete information, and for VSLs demonstrate to drivers there is a reason to slow down, as in school zones where drivers see the kids, make the VSL zone short where drivers see the workers in an active school zone. 13. Has your agency given any consideration to in-vehicle or smartphone application messages for deploying ATM in a connected vehicle environment? Has your agency considered phasing out the deployment or operations of traditional ITS or on-road signing, such as DMS? If yes, please explain. Blaine Leonard is the person to talk to about connected vehicle work in Utah – still taking baby steps to make sure the current technologies like VSLs work first before doing more with them. Virginia respondent – Mike Fontaine, michael.fontaine@vdot.virginia.gov 1. What ATM strategies are currently deployed or being considered for deployment in your jurisdiction? Permanent deployments: • VSLs – I-77 – primary for inclement weather: dense fog; regulatory; very rural with a lot of thru traffic, use VSLs for non-recurring events, incident management, or work zones • Dynamic Shoulder Lanes, Advisory VSL, Queue Warning – I-66 from US 29 in Gainesville to I-495 Beltway – begin operations in 2015, intent to be permanent, but will be removed for highway widening for I-66 Express Lanes project, not sure if will re-deploy.

123 Intended for congestion management; replaced the single lane signage of the time-of-day shoulder use system on I-66 that had been operational for many years to make it fully dynamic, open during weekends and off-peak periods, as needed, along the same shoulder running stretch as before from US 50 to I-495 Beltway. As part of a construction project for that, we will add a new Express Lane in each direction; all of the ATM gantries will come down later this year, and may or may not be replaced at the end of the project. Temporary deployments: • Queue warning – planning these for smart work zone deployments this year • VSLs – operated 10 years ago on I-495 for Woodrow Wilson Bridge, but had challenges with visibility and enforcement Also have, though not truly dynamic ATM systems: • Lane Reversal – I-95 Express Lanes – a pre-defined schedule that rarely changes for real- time conditions • VSL - Hampton Roads tunnels, technically are dynamic speed limits, but it is manually controlled from the tunnel control center, very reactive and operator judgment 2. For each ATM strategy, please specify how is ATM information disseminated to drivers, i.e., overhead signs (one sign for each lane), overhead signs (one sign for multiple lanes), lateral roadside signs, in-vehicle displays, smartphone applications, other. • I-77: roadside signs on both left and right • I-66: overhead signs, one sign for each lane 3. For each ATM strategy, what kind of signage is used? i.e., fully dynamic sign, static sign with dynamic elements, other. • I-77: some fully dynamic – full matrix, full-color signs to deploy other messages; others are static speed limit sign with dynamic numerals (see response to question 6) • I-66: fully dynamic: full matrix, full color 4. What factors were used to prioritize funding for your ATM deployments? i.e., safety benefits, mobility benefits, other. • I-77 – safety • I-66 – mobility, have seen safety benefits also, but very tied to congestion reduction on the corridor where there is the dynamic shoulder from US 50 to the I-495 Beltway – biggest benefits in this segment • Queue warning in work zones for safety 5. What resources did your agency use (e.g., documents, guidance, peer deployments) to inform the design of your ATM deployment, including signage, display format of information, content, timing, and priority for displaying information, as applicable? Peer experiences were very significant, particularly for I-66, and included many discussions with MnDOT and WSDOT. For I-77, there was a prolonged study phase on how to develop the algorithm and sign configurations, which included a lot of outreach to Utah, WSDOT,

124 and others by the contractor (Kimley-Horn); a guidance document by FHWA on Weather VSLs was also used. 6. How did your agency choose the types and distribution of signs (e.g., fully dynamic over every lane at ¼-mile spacing) for your deployment? • I-66 generally was planned as earlier deployments (Seattle, Minneapolis) and had ½ mile spacing and lane-by-lane signs built in from the start of the project. Looking ahead for similar deployments elsewhere, there might be some resistance on that sort of rigid design in such an infrastructure-heavy deployment, to make it more cost driven. • I-77 was partially designed to a budget – ideally, we would want full matrix DMS on all signs, but offset some costs with the cutout signs – this approach is a matter of asking how can we accomplish what we want without compromising system effectiveness? 7. Has your agency made any changes to the way information is presented to drivers since your ATM deployment first began operations? (e.g., different message wording, different symbols, different spacing or sign configurations, different types of signage, distance presented upstream, different thresholds in algorithms for operations, etc.) If yes, please explain. • I-66 – September 2015 – realized very quickly that the posted speed limits were abnormally low; the advisory VSL system was taken offline for 3 months reactivated with updated algorithm. The lane control signs have a request to experiment with yellow merge arrows; began using the yellow merge arrows at the on-ramps when the dynamic shoulder is closed in November 2016 instead of just the red X. We have seen that drivers are not responding to the VSL until they see a need to slow down – it is a function more of congestion than the sign display. • I-77 – developed an algorithm based on driver behavior prior to activation, then examine whether behavior has shifted to slower speeds after activation, and it has. Now looking to re-write algorithm as part of an iterative process: as we see driver behavior fall in line with what we want, we can adjust the algorithm to slow traffic to lower speeds that are even closer to the “optimum” safe speed. 8. How does your agency balance the needs of drivers versus the infrastructure, operations, and maintenance costs for ATM? Take it on a case-by-case basis – staff at the chief engineer level are more concerned with installing infrastructure that’ll have to be maintained knowing that in the next 10-20 years with connected vehicles we may not need to provide that infrastructure; asking: can we do something now, but not overbuild? Striking this balance and answering these questions is not a formal process, but something that comes through the systems engineering process and concept of operations. 9. Has your agency evaluated the return on investment of your ATM deployments? If so, how was this done? VDOT conducted an evaluation of the I-66 system impacts. Found no significant improvement in crashes on segments without dynamic shoulder use: the shoulder lane is open almost double the hours it was before. VTRC report attempted to monetize the I-66

125 benefits on I-66 – primarily focused on travel time savings; didn’t look at operations and maintenance. 10. If you could do the ATM deployment again, what would you do differently and why? Specifically, would you use the same method for displaying information to drivers (e.g., the types and distribution of signs: fully dynamic over every lane at ¼-mile spacing)? • I-77 generally has been pretty successful – there was a well-defined need, and drivers are on-board. Relatively limited data before deployment: a “before study” was conducted and a lot of decisions were made based on short-term studies. It would’ve been nice to have better data so we could get the algorithm better initially, we can now iterate the algorithm because a lot more sensors are collecting a lot better data; we weren’t far off with the initial algorithm, but it would’ve been nice to have it a little better at first. • I-66 – it was a challenge to set expectations. The system is producing no impact during peak periods because the shoulder lane was already open to travel, so there is no capacity improvement, the major impacts are in the off-peak periods when the dynamic shoulder is now open longer than before. Thus, the average commuter does not see those benefits – I- 66 is massively over capacity and the ATM system hasn’t changed that. Fundamentally, it’s the added capacity on the shoulder that brings in the benefit. 11. Has your agency surveyed or received feedback from drivers, either positive or negative, about the type of information they receive or the dissemination mechanism that is used? Please explain what drivers do and do not like about the deployment. • A local delegate in Northern Virginia does a regular survey of constituents, which recently included questions related to the overall value of the ATM system, not about user impacts; • There was a driver survey for the diagonal arrow – TRB paper on the effectiveness was e- mailed to Jeremy. 12. What types of information and dissemination mechanism do you believe is most effective for ATM? Varies for the need and location: • I-77 – relatively very easy, with an initial VMS to announce why a lower speed is needed, e.g., “Fog Ahead,” then we see good compliance; • I-66 – drivers are looking for travel time info (provided on VMS separate from the ATM system), the nature of the delay, and what lanes are closed; generally speaking, the VSLs are publicly viewed as the least valuable part of the system, partially because drivers see them as redundant with other items, i.e., travel time; drivers on I-66 expect it to be congested. 13. Has your agency given any consideration to in-vehicle or smartphone application messages for deploying ATM in a connected vehicle environment? Has your agency considered phasing out the deployment or operations of traditional ITS or on-road signing, such as DMS? If yes, please explain. • This question is an active and hot topic for discussion: what’s the future of traveler information and to what degree do we have to rely on DMS that we have to build and maintain vs. other providers and applications like WAZE. The VDOT traveler information program is going through a lot of strategic planning to determine the future of the group:

126 “the future of transportation signing in Virginia,” and asking what would it look like to move to a virtual signing system? • Starting to get more pushback from administrators to justify investments in deploying new traditional ITS. The deployed ITS is fulfilling a critical role now, but what does it look like in the future? • Currently conducting research to ask how should we provide this information, what can we do now that we couldn’t before? • VDOT is working with the VTTI to develop a smartphone app to replicate the I-66 ATM signs on the phone and replicate in-vehicle messaging. The app is not publicly available but is currently being tested with 250 people to examine its utility in showing a virtual version of the gantry signs. We are also testing what we cannot do on the signs now: testing work zone management through the app, e.g., work vehicle trying to enter the freeway, the app issues a truck entering warning. • VDOT is working with UVa to examine virtual DMS – geofence gateways to push information to the driver. • How do we work with the third-party providers to push information to them? What’s the role of government vs. private sectors? WAZE exchange more about the 511 feed with incidents. In future considering providing a central clearinghouse with public data, instead of separate data feeds. Washington respondent – Vinh Dang, dangv@wsdot.wa.gov 1. What ATM strategies are currently deployed or being considered for deployment in your jurisdiction? All permanent deployments: • VSLs – rural areas: weather correlated with surface condition – done on a manual basis; • VSLs – urban areas: all based on speed and occupancy, and queue detection – information presented is retained by drivers because it is presented ~1 minute upstream, otherwise can disregard and forget. Use yellow caution arrow on outside lane at merge points when traffic volume from on-ramp is heavy. • Dynamic lane reversal – on I-5 and I-90, generally based on time of day – people count on it based on the hours they know it’s open and see people rush in before it closes to change directions – sometimes adjust the times based on incidents, sometimes open early; 90 express lanes – just finished the scheduling for special events, but these permanently closing soon for light rail; • Dynamic lane use; • Dynamic shoulder use – open on 405 now, being considered on I-5 - use three thresholds on the shoulder – occupancy, speed, and volume. • 405 express lanes – being extended south, algorithm is based on congestion on both GP and toll lanes; • ICM – work with City of Seattle to do a small ICM strategy for I-5 North into downtown 2. For each ATM strategy, please specify how is ATM information disseminated to drivers, i.e., overhead signs (one sign for each lane), overhead signs (one sign for multiple lanes), lateral roadside signs, in-vehicle displays, smartphone applications, other.

127 Signs over each lane on I-5, I-90, SR 520 in Seattle; rural VSLs mostly on single overhead DMS; most overhead; the interim VSLs on I-90 had signs on the side. New dynamic shoulder deployment is “ATM lite” with display only over shoulder. 3. For each ATM strategy, what kind of signage is used? i.e., fully dynamic sign, static sign with dynamic elements, other. Most deployments are fully dynamic, full color; rural VSLs are fully dynamic but not full color; some static signs with dynamic elements for travel times. 4. What factors were used to prioritize funding for your ATM deployments? i.e., safety benefits, mobility benefits, other. Safety – has been the primary factor, lane control signage allows drivers to merge with traffic at slower speeds rather than pull all the way up and get trapped; look also for the potential of conflicts, not just the crashes themselves – look at number of evasive maneuvers and brake lights; weather. Mobility – was the primary factor for shoulder lanes, creates increased capacity 5. What resources did your agency use (e.g., documents, guidance, peer deployments) to inform the design of your ATM deployment, including signage, display format of information, content, timing, and priority for displaying information, as applicable? Conduct research to find peers doing similar things; follow systems engineering process to identify the problem: bottleneck, crashes 6. How did your agency choose the types and distribution of signs (e.g., fully dynamic over every lane at ¼-mile spacing) for your deployment? The system was designed based on a free-flow speed of 60 mph, i.e., if we detect an incident and want to close the lane, how much advanced notice is needed for the driver to perceive and react – determined to be about 30 seconds, which is a ½-mile. They also use ½-mile spacing in Europe. Rural VSLs are located downstream of interchanges – a big reason to not place them between interchanges is the lack of power; this placement is logical based on plowing operations, which is done from one interchange to the next and will be about the same for the entire segment, e.g., spacing is up to 3-4 miles on I-90 rural VSL deployment at Snoqualmie. 7. Has your agency made any changes to the way information is presented to drivers since your ATM deployment first began operations? (e.g., different message wording, different symbols, different spacing or sign configurations, different types of signage, distance presented upstream, different thresholds in algorithms for operations, etc.) If yes, please explain. Reduced minimum speed that was displayed on VSLs to 30 mph, and made other algorithm adjustments to smooth operations. Roadside infrastructure is costly – heading toward sending that information to the vehicle, and now do more “ATM lite” to reduce infrastructure needs, where signs are not over each lane. Future are focused on shoulder use to increase capacity, e.g., I-405 shoulder lane just opened, I-90 from Issaquah to Bellevue coming by 2020 with shoulder.

128 When the in-vehicle technology is present, considering HOV system conversion to AV lanes where vehicles have short gaps between them to increase capacity. For future shoulder deployments, we are looking to be able to use a facility at multiple speeds rather than just when it is slower; we can do this by designing the shoulder with a slightly wider buffer, and apply strategies that we use when there is planned activity for work zone, e.g., during times not have a shoulder, we beef up the IRT program to be more responsive to stalled vehicles. In our messaging and signage, we have learned when to use action verbs – used to say, “shoulder use permitted” now tell people “use shoulder when metered.” For ramp metering, increased storage by posting sign “form two lines when metered” 8. How does your agency balance the needs of drivers versus the infrastructure, operations, and maintenance costs for ATM? All the deployments are one-way communications so we do not really know driver needs; we are looking for better ways to get driver feedback to complete the circle – we can look at traffic data and see that people respond to the messages, but interested in dynamic routing and would be helpful to know whether people respond to alternate route suggestions. 9. Has your agency evaluated the return on investment of your ATM deployments? If so, how was this done? We looked at crash history before/after, but have not pinpointed a number of prevented crashes. For the shoulder project, we will have a performance report on mobility improvements. 10. If you could do the ATM deployment again, what would you do differently and why? Specifically, would you use the same method for displaying information to drivers (e.g., the types and distribution of signs: fully dynamic over every lane at ¼-mile spacing)? We have learned to keep it short and sweet – the new shoulder use signs are small – drivers do not have time to read a lot of information. 11. Has your agency surveyed or received feedback from drivers, either positive or negative, about the type of information they receive or the dissemination mechanism that is used? Please explain what drivers do and do not like about the deployment. We informed the public about how the VSL system displayed a speed limit, not the actual speed – adjusted down to 30 mph – media and WSDOT educate public. No feedback about the symbols – we have received some pushback regarding MUTCD compliance of symbols for merge or caution arrow, but the pieces are consistent with other applications in the MUTCD, and driver reactions anecdotally convey understanding. Anecdotally, no confusion with which direction or lane needs to merge – the message commands attention and is not confusing, drivers generally follow the merge arrows when they see them depending on their speed, and use of the signs does not increase crashes. The yellow caution arrow that is used for slow traffic ahead, junction warning, and a queue warning when there is heavy merge traffic is also controversial.

129 12. What types of information and dissemination mechanism do you believe is most effective for ATM? All of the above – it all depends on the specific conditions and location types. Have noticed it is more effective to target your audience in a message in addition to route information – “Bellevue traffic…,” “Downtown traffic…” to help drivers digest and filter the information that is needed – and not have to read all of it. 13. Has your agency given any consideration to in-vehicle or smartphone application messages for deploying ATM in a connected vehicle environment? Has your agency considered phasing out the deployment or operations of traditional ITS or on-road signing, such as DMS? If yes, please explain. Mature enough with traditional ITS in metro area for most part, but still deploying traditional ITS as needed; Express Lanes do not have a lot of ITS. We can make up for the lack of DMS by using social media to alert drivers of incidents, especially in rural areas; distracted driving is discouraged; have had positive feedback from this approach. Have noticed that WAZE sometimes has stale info. Wyoming respondent – Vince Garcia, vince.garcia@wyo.gov 1. What ATM strategies are currently deployed or being considered for deployment in your jurisdiction? Permanent VSLs – primarily for weather, have also used the permanent infrastructure for reduced work zone speed limits. Note also – connected vehicle pilot applications. 2. For each ATM strategy, please specify how is ATM information disseminated to drivers, i.e., overhead signs (one sign for each lane), overhead signs (one sign for multiple lanes), lateral roadside signs, in-vehicle displays, smartphone applications, other. Roadside signs: for Interstate, have one sign on each side of the roadway; for 2-lane highways only have one side on the right. 3. For each ATM strategy, what kind of signage is used? i.e., fully dynamic sign, static sign with dynamic elements, other. All VSL signs are static signs with dynamic amber on all black characters – have used the same sign design (with different vendors) since the first deployment. Before deployment, a visibility study determined this color scheme to be most effective to read in all conditions, including blowing snow conditions. The white LED lights may have improved since then for new, full-color matrix signs, but this – understand there could be newer full color that may be better, white LED this was previously found to be less effective. 4. What factors were used to prioritize funding for your ATM deployments? i.e., safety benefits, mobility benefits, other. Safety benefits primarily; we have seen secondary mobility benefits because we are able to reduce the number of closures by reducing crashes

130 5. What resources did your agency use (e.g., documents, guidance, peer deployments) to inform the design of your ATM deployment, including signage, display format of information, content, timing, and priority for displaying information, as applicable? The University of Wyoming conducted a lot of research for the VSL system – the first deployment was done as a research project, and they helped the DOT develop an extensive algorithm for use by the TMC. This algorithm has continued to be fine-tuned and enhanced; snowplow drivers are now able to provide inputs for reducing or raising the VSL, e.g. Essentially, the VSL system was designed without outside assistance because Wyoming was one of the first to deploy such a system. Wyoming has and continues to conduct peer exchanges with other states – Utah, Colorado, South Dakota, Oregon. 6. How did your agency choose the types and distribution of signs (e.g., fully dynamic over every lane at ¼-mile spacing) for your deployment? Initially deployed mostly just at on-ramps. We installed signs in between for long segments when problem areas were identified, specifically weather problem spots. For example, if VSL signs at on-ramps were 10 miles apart and noticed a major recurring weather issue at mile six, we would have to use the sign at the first on-ramp to reduce the speed, but drivers would lose trust in the system with a reduced speed limit for that long with no visible or experienced issue, so we densified the system for those locations. The VSL signs are supplemented by full DMS to post a reason for reduced speeds, there were existing DMS in most cases, but also added some to get additional info to drivers. 7. Has your agency made any changes to the way information is presented to drivers since your ATM deployment first began operations? (e.g., different message wording, different symbols, different spacing or sign configurations, different types of signage, distance presented upstream, different thresholds in algorithms for operations, etc.) If yes, please explain. Algorithm continues to be updated. We use AC power wherever possible, but had to install DC power a few times when could not install AC in a cost- or time-effective manner. Some locations can go weeks where don’t get sunlight for solar power, so we are beginning to add wind power at those locations. To expand the benefits seen by the VSL systems, we are conducting a study on displaying advisory speeds on existing DMS. We are expecting to display these advisory speeds on the DMS in a year, following additional study. 8. How does your agency balance the needs of drivers versus the infrastructure, operations, and maintenance costs for ATM? We have developed crash mitigation factors to see if it makes sense to deploy VSLs; we do not believe VSLs should be installed everywhere, so we are deploying strategically where it makes sense to do so. 9. Has your agency evaluated the return on investment of your ATM deployments? If so, how was this done? Yes, the study showed that the return on investment is within a year. Regarding the system, vendors are required to provide signs that last 20 years.

131 10. If you could do the ATM deployment again, what would you do differently and why? Specifically, would you use the same method for displaying information to drivers (e.g., the types and distribution of signs: fully dynamic over every lane at ¼-mile spacing)? We should have deployed the VSL systems sooner, possibly before or instead of some DMS installations – the reduced speed on the VSL signage provides the drivers what they need to know, although the DMS are nice to have. Our first VSL installation began in 2009; had several large multi-vehicle crashes that spurned us to do them, but we could have been more proactive. 11. Has your agency surveyed or received feedback from drivers, either positive or negative, about the type of information they receive or the dissemination mechanism that is used? Please explain what drivers do and do not like about the deployment. No formal surveys. Receive frequent anecdotal feedback that the VSLs are very helpful; people comment frequently that they are effective. Even if we raise the speed limit, we get told when they think it is higher than it should be. 12. What types of information and dissemination mechanism do you believe is most effective for ATM? The VSL display provides drivers what they need to know and is most important, but it’s nice to have the DMS too. 13. Has your agency given any consideration to in-vehicle or smartphone application messages for deploying ATM in a connected vehicle environment? Has your agency considered phasing out the deployment or operations of traditional ITS or on-road signing, such as DMS? If yes, please explain. Planning to incorporate in-vehicle messaging as a part of the CV Pilot project, and it will not just be DSRC based. Partnering with private-sector providers of in-vehicle systems for commercial vehicles, also with Sirius XM radio – would provide drivers with speed limits and general traffic information based on their location. We are still in the planning and design stages and cannot guarantee what the third-party will do with the information we provide them, but are exploring the provision of both visual and audio messages to drivers. Continuing to deploy infrastructure, as needed. Links to Referenced Sources: 1. http://www.fdot.gov/research/Completed_Proj/Summary_CN/FDOT_BD548-24_rpt.pdf 1. International Road Dynamics. (2005). Dynamic Lane Merge Systems. MdSHA. 1. https://international.fhwa.dot.gov/pubs/pl07012/atm_eu07.pdf 1. https://international.fhwa.dot.gov/pubs/pl11004/pl11004.pdf 1. https://ops.fhwa.dot.gov/publications/fhwahop10031/fhwahop10031.pdf 1. http://www.traffictechnologytoday.com/news.php?NewsID=83107 1. https://itunes.apple.com/us/app/idaho-511/id842945399?mt=8# 1. https://itunes.apple.com/us/app/louisiana-511/id1141141943?mt=8 1. https://itunes.apple.com/us/app/minnesota-511/id664709236?mt=8 1. https://itunes.apple.com/us/app/iowa-511/id528047799?mt=8

132 1. https://ntl.bts.gov/lib/57000/57000/57027/Rewrite_Final_1.13.16_FHWA-JPO-15-232.pdf 1. https://www.its.dot.gov/research_archives/dma/bundle/inflo_plan.htm 1. https://ntl.bts.gov/lib/56000/56200/56240/FHWA-JPO-15-223.pdf Follow-Up Survey Results Illinois State Toll Highway Authority – Elyse Morgan What ATM strategies are currently deployed or being considered for deployment in your jurisdiction? Select all that apply. Dynamic Lane Use Control Permanent Deployment Dynamic Merge Control Under Consideration - Permanent Deployment Dynamic Shoulder Lanes Permanent Deployment Queue Warning Permanent Deployment VSL Under Consideration - Permanent Deployment For each ATM strategy, how is ATM information disseminated to drivers? Select all that apply. If n/a, leave blank. Dynamic Lane Use Control Overhead Signs (one sign for each lane) Dynamic Merge Control Overhead Signs (one sign for multiple lanes) Dynamic Shoulder Lanes Overhead Signs (one sign for each lane) Queue Warning Overhead Signs (one sign for multiple lanes), Lateral Roadside Signs VSL Overhead Signs (one sign for each lane) For each ATM strategy, what kind of signage is used? Select all that apply. If n/a, leave blank. Dynamic Lane Use Control Fully dynamic sign Dynamic Merge Control Fully dynamic sign Dynamic Shoulder Lanes Fully dynamic sign Queue Warning Fully dynamic sign, Static sign with dynamic elements VSL Fully dynamic sign Other: For queue warning, we use both DMS and static signs with flashing beacons What measures and data sources did your agency use to project the following when considering deployment of ATM strategies? Select all that apply. If n/a, leave blank. Dynamic Lane Use Control Costs, Driver needs, Mobility impacts, Safety impacts Dynamic Merge Control Mobility impacts, Safety impacts Dynamic Shoulder Lanes Mobility impacts, Safety impacts Queue Warning Mobility impacts, Safety impacts VSL Safety impacts What resources did your agency use to inform the design and operations of your ATM deployment, including signage, display format of information, content, timing, and priority for displaying information, as applicable? Please list relevant documentation, peer deployments, or other resources used. Published Guidance and Reports: FHWA studies on ATM, MUTCD

133 Interactions and Sample Documentation from Peer Deployments: Visits and discussions with Washington DOT and Virginia DOT In what topic areas did you feel that guidelines and resources were lacking and would have been helpful for planning the design and operations of the ATM deployment? Select all that apply. Message display (graphics vs. text) What types of guidelines and resources, if any, would have pre-empted or reduced the need for your agency to conduct peer exchanges with other ATM deploying agencies? Respondent skipped this question Has your agency evaluated the safety or mobility benefits of your ATM deployments? Yes Please provide any additional information you would like to share regarding ATM initiatives and deployments. There needs to be better guidance and information about lane control sign placement and the type of graphics that can be used on the signs. This caused a lot of back and forth and multiple design and operational strategy changes. Metropolitan Washington Council of Governments – Andrew Meese Staff at the Metropolitan Washington Council of Governments did not proceed with filling out the survey for the following provided reasons: • Our MPO has not taken an active role in the consideration and deployment of ATM systems/equipment. For example, we cannot speak to the reasons why any particular ATM strategy was deployed, nor what resources were used. • While we have anecdotes, we do not have an inventory of systems deployed in our region, nor of their specific features and operational characteristics. • Overall, we as an MPO generally remain at the strategic and long-range planning level, and do not delve into operational characteristics such as those addressed by the survey. Michigan Respondent – Jennifer Foley What ATM strategies are currently deployed or being considered for deployment in your jurisdiction? Select all that apply. Dynamic Lane Use Control Permanent Deployment Dynamic Merge Control Permanent Deployment Dynamic Shoulder Lanes Permanent Deployment Queue Warning Permanent Deployment VSL Permanent Deployment Other: Currently reviewing HOV and HOT alternatives For each ATM strategy, how is ATM information disseminated to drivers? Select all that apply. If n/a, leave blank. Dynamic Lane Use Control Overhead Signs (one sign for each lane)

134 Dynamic Merge Control Overhead Signs (one sign for each lane) Dynamic Shoulder Lanes Overhead Signs (one sign for each lane) Queue Warning Overhead Signs (one sign for each lane) VSL Overhead Signs (one sign for each lane) For each ATM strategy, what kind of signage is used? Select all that apply. If n/a, leave blank. Dynamic Lane Use Control Fully dynamic sign Dynamic Merge Control Fully dynamic sign Dynamic Shoulder Lanes Fully dynamic sign Queue Warning Fully dynamic sign VSL Fully dynamic sign Other: There are additional DMS in addition to the signs over each lane for additional information. What measures and data sources did your agency use to project the following when considering deployment of ATM strategies? Select all that apply. If n/a, leave blank. Dynamic Lane Use Control Costs, Driver needs, Mobility impacts, Safety impacts Dynamic Merge Control Costs, Driver needs, Mobility impacts, Safety impacts Dynamic Shoulder Lanes Costs, Driver needs, Mobility impacts, Safety impacts Queue Warning Costs, Driver needs, Mobility impacts, Safety impacts VSL Costs, Driver needs, Mobility impacts, Safety impacts What resources did your agency use to inform the design and operations of your ATM deployment, including signage, display format of information, content, timing, and priority for displaying information, as applicable? Please list relevant documentation, peer deployments, or other resources used. Published Guidance and Reports: At the time of deployment there was little guidance, we researched other State deployments Interactions and Sample Documentation from Peer Deployments: WSDOT, MnDOT, VDOT, CDOT, United Kingdom Other Resources: FHWA TOPS BC Tool In what topic areas did you feel that guidelines and resources were lacking and would have been helpful for planning the design and operations of the ATM deployment? Select all that apply. Sign placement (spacing, overhead vs. side-mount) Message display (graphics vs. text) Software and algorithm development for automated operations Priority of message to display Other: All of these are important, it varies depending on location and need. The best information is what other states have and how they use it to see if it would apply to your own application and what they learned. What types of guidelines and resources, if any, would have pre-empted or reduced the need for your agency to conduct peer exchanges with other ATM deploying agencies? State of the Practice reports

135 Has your agency evaluated the safety or mobility benefits of your ATM deployments? Yes Please provide any additional information you would like to share regarding ATM initiatives and deployments. It is important for states to understand the maintenance and communication costs of each deployment. MDOT is starting to collect this data along with a research study regarding the safety impacts of ATM. Minnesota Respondent – Brian Kary What measures and data sources did your agency use to project the following when considering deployment of ATM strategies? Select all that apply. If n/a, leave blank. Dynamic Lane Use Control Mobility impacts, Safety impacts Dynamic Shoulder Lanes Mobility impacts, Safety impacts VSL Mobility impacts, Safety impacts What resources did your agency use to inform the design and operations of your ATM deployment, including signage, display format of information, content, timing, and priority for displaying information, as applicable? Please list relevant documentation, peer deployments, or other resources used. Published Guidance and Reports: European Scan Tour on Hard Shoulder Use and ATM Interactions and Sample Documentation from Peer Deployments: Review of European deployments, Interaction with United Kingdom staff In what topic areas did you feel that guidelines and resources were lacking and would have been helpful for planning the design and operations of the ATM deployment? Select all that apply. Other (please specify): We pre-dated anything in the US so were following European examples. Collaborated with Washington State DOT in developing concepts for US. What types of guidelines and resources, if any, would have pre-empted or reduced the need for your agency to conduct peer exchanges with other ATM deploying agencies? Guidelines on sign spacing, message display Has your agency evaluated the safety or mobility benefits of your ATM deployments? Yes North Central Texas Council of Governments Respondent – Natalie Bettger What ATM strategies are currently deployed or being considered for deployment in your jurisdiction? Select all that apply. Dynamic Lane Reversal Permanent Deployment Dynamic Lane Use Control Permanent Deployment Dynamic Shoulder Lanes Temporary Deployment (e.g., work zones, short term until roadway widening)

136 For each ATM strategy, how is ATM information disseminated to drivers? Select all that apply. If n/a, leave blank. Dynamic Lane Reversal Overhead Signs (one sign for multiple lanes), Lateral Roadside Signs Dynamic Lane Use Control Overhead Signs (one sign for each lane) Dynamic Shoulder Lanes Overhead Signs (one sign for each lane), Lateral Roadside Signs For each ATM strategy, what kind of signage is used? Select all that apply. If n/a, leave blank. Dynamic Lane Reversal Static sign with dynamic elements Dynamic Lane Use Control Fully dynamic sign Dynamic Shoulder Lanes Static sign with dynamic elements What measures and data sources did your agency use to project the following when considering deployment of ATM strategies? Select all that apply. If n/a, leave blank. Dynamic Lane Reversal Costs, Driver needs, Mobility impacts, Safety impacts Dynamic Lane Use Control Costs, Driver needs, Mobility impacts, Safety impacts Dynamic Shoulder Lanes Costs, Driver needs, Mobility impacts What resources did your agency use to inform the design and operations of your ATM deployment, including signage, display format of information, content, timing, and priority for displaying information, as applicable? Please list relevant documentation, peer deployments, or other resources used. Published Guidance and Reports: MUTCD In what topic areas did you feel that guidelines and resources were lacking and would have been helpful for planning the design and operations of the ATM deployment? Select all that apply. Message display (graphics vs. text), Software and algorithm development for automated operations What types of guidelines and resources, if any, would have pre-empted or reduced the need for your agency to conduct peer exchanges with other ATM deploying agencies? Best practices in standard operating procedures for ATM deployment operations. Partners to engage in ATM deployment development, i.e. police, fire, tow truck operators, etc. Has your agency evaluated the safety or mobility benefits of your ATM deployments? Yes Please provide any additional information you would like to share regarding ATM initiatives and deployments. Respondent skipped this question Oregon Respondent – Galen McGill What ATM strategies are currently deployed or being considered for deployment in your jurisdiction? Select all that apply. Dynamic Shoulder Lanes Under Consideration - Permanent Deployment Queue Warning Permanent Deployment

137 VSL Permanent Deployment For each ATM strategy, how is ATM information disseminated to drivers? Select all that apply. If n/a, leave blank. Queue Warning Signage varies for different deployments Other VSL Signage varies for different deployments, Other (please specify other ways ATM information is disseminated to drivers and the applicable ATM strategy) Data is available through data portal. For each ATM strategy, what kind of signage is used? Select all that apply. If n/a, leave blank. Queue Warning Fully dynamic sign VSL Fully dynamic sign, Static sign with dynamic elements What measures and data sources did your agency use to project the following when considering deployment of ATM strategies? Select all that apply. If n/a, leave blank. Queue Warning Costs, Mobility impacts, Safety impacts VSL Costs, Mobility impacts, Safety impacts What resources did your agency use to inform the design and operations of your ATM deployment, including signage, display format of information, content, timing, and priority for displaying information, as applicable? Please list relevant documentation, peer deployments, or other resources used. Published Guidance and Reports: International Scan report Interactions and Sample Documentation from Peer Deployments: Input/lessons learned from peer states. FHWA funded peer exchange meeting. In what topic areas did you feel that guidelines and resources were lacking and would have been helpful for planning the design and operations of the ATM deployment? Select all that apply. Sign placement (spacing, overhead vs. side-mount), Software and algorithm development for automated operations, Priority of message to display What types of guidelines and resources, if any, would have pre-empted or reduced the need for your agency to conduct peer exchanges with other ATM deploying agencies? Case studies documenting peer experiences, but it is hard to replace the value of talking directly with someone else that has experience with deployment. Has your agency evaluated the safety or mobility benefits of your ATM deployments? Yes Please provide any additional information you would like to share regarding ATM initiatives and deployments. Respondent skipped this question

138 Virginia Respondent – Mike Fontaine What measures and data sources did your agency use to project the following when considering deployment of ATM strategies? Select all that apply. If n/a, leave blank. Dynamic Lane Use Control Costs, Driver needs, Mobility impacts, Safety impacts Dynamic Shoulder Lanes Costs, Driver needs, Mobility impacts, Safety impacts Queue Warning Driver needs, Safety impacts VSL Costs, Driver needs, Mobility impacts, Safety impacts What resources did your agency use to inform the design and operations of your ATM deployment, including signage, display format of information, content, timing, and priority for displaying information, as applicable? Please list relevant documentation, peer deployments, or other resources used. Published Guidance and Reports: For I-77, FHWA guidance on VSL for wet weather (https://safety.fhwa.dot.gov/speedmgt/ref_mats/fhwasa12022/fhwasa12022.pdf) Interactions and Sample Documentation from Peer Deployments: Peer discussions/interviews with Washington state and Minnesota for I-66; Oregon, Utah, and Tennessee for I-77 In what topic areas did you feel that guidelines and resources were lacking and would have been helpful for planning the design and operations of the ATM deployment? Select all that apply. Message display (graphics vs. text), Software and algorithm development for automated operations What types of guidelines and resources, if any, would have pre-empted or reduced the need for your agency to conduct peer exchanges with other ATM deploying agencies? While effectiveness data is widely available, sometimes there is a lack of data on exactly how systems were operated. Since that influences effectiveness, more data on algorithms would be very useful. Has your agency evaluated the safety or mobility benefits of your ATM deployments? Yes Please provide any additional information you would like to share regarding ATM initiatives and deployments. The I-66 ATM was deactivated in Spring 2018 due to Express Lane construction. The final evaluation report can be found here: http://www.virginiadot.org/vtrc/main/online_reports/pdf/19-r7.pdf I-77 weather VSLs are still in place. The evaluation report can be found here: http://www.virginiadot.org/vtrc/main/online_reports/pdf/19-r6.pdf