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.
Task Analysis of Driver Merging Behavior at Freeway Entrance Ramps . . . . . . . . . . . . . . . .12-2 Reducing Wrong-Way Entries onto Freeway Exit Ramps . . . . . . . . . . . . . . . . . . . . . . . . . . . .12-4 Driver Expectations at Freeway Lane Drops and Lane Reductions . . . . . . . . . . . . . . . . . . . . .12-6 Driver Information Needs at Complex Interchanges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .12-8 Arrow-per-Lane Sign Design to Support Driver Navigation . . . . . . . . . . . . . . . . . . . . . . . . .12-10 Driver Behavioral Trends Based on Exit Ramp Geometry . . . . . . . . . . . . . . . . . . . . . . . . . . .12-12 12-1 C H A P T E R 12 Interchanges
T AS K A NA LY SI S OF D RI VE R M ER GI NG B EH AV IO R AT F RE EW AY E NT RA NC E R AM PS In tr od uc ti on Merging fro m an entrance ramp onto the freeway ma inline can be a challenging task for drivers. For exa mp le, 36% of all ra mp accidents on urban interstates in Northern Virginia occurred when drivers were entering the freeway ( 1 ). In addition, simulator data showed that, when me rging, drivers move thei r hands to positions in which they can exert mo re vehicle control ( 2 ). Although the driver task can be broken up into a series of subtasks, the process is mo re dyna mi c than mechanistic: both ma inline and me rging drivers can detect the gaps available and decide to change their speed accordingly. De si gn Gu id e lin es Task Distance/Time Derivation Driver and Roadway Factors 1. Initial steering component : Drivers steer to transition from the entrance ramp to the Speed- Change Lane (S CL). ⢠Constant time of approximately 1 s, derived from research. NA 2. Acceleratio n : Drivers accelerate to obtain an unobstructed view of mainline freewa y traff ic. ⢠Time determined by the travel distance required to see approaching vehicles on the mainline. ⢠The observed 85 th percentile maximum comfortable acceleration is 2.0 m/s 2 ( 3 ). ⢠Affected by the controlling ramp curvature ( 4 ). 3. Gap search : After seeing the ramp nose, drivers begin to search for a gap in the mainline traffic to use for their merge. ⢠0.25 to 0.5 s is required to detect the angular velocity of the lag vehicle on the freeway mainline. ⢠No stable base for judging speed/position of freeway vehicles ( 4 ). ⢠Drivers who force in accept smaller gaps and accelerate ( 5 ). ⢠With heavy congestion, zip merging at ramp end happens instead of gap search ( 6 ). 4. Merge steering : Drivers steer to transition from the SCL to the freeway mainline. ⢠85% of observed vehicles merge comfortably in 375 m ( 7 ). ⢠SCL lengths over 425 m do not improve merging behavior ( 7 ). On tapered ramps vs. parallel ramps: ⢠A greater ramp length is used to merge ( 8 ). ⢠Drivers merge mo re aggressively ( 9 ). 5. Abort : Drivers who do not find a gap to merge into decelerate to a stop before running out of SCL. ⢠Time determined by the angular velocity of the approaching ramp end. NA S UB TA SK L OC AT IO NS Source: adapted from Aha mmed et al . ( 3 ) Ba sed Primarily on Ex pert Judgment Based Equally on Expert Judgment and Empirical Dat a Based Primarily on Em pirical Da ta 1 2 3 4 5 SC L La ne 1 La ne 2 HFG INTERCHANGES Version 2.0 12-2
Di scu ssi on Initial steering component : The initial steering co mp onent occurs when drivers transition from the entrance ra mp onto the SCL. This steering ti me is approximately 1 s in length, derived from empirical research on steering ( 4 ). Acceleration component: During the acceleration component, drivers accelerate to obtain a view of the ma inline traffic on the freeway. This acceleration is controlled by the ra mp curvature ( 4 ). Additionally, drivers cannot begin the next component, gap search, until they have an unobstructed vie w of the ma inline traffic. Hunter and Machemehl ( 7 ) found that ra mp s with adequate sight distance and SCL lengths le d to small acceleration levels, while ra mp s without adequate sight distance and SCL lengths caused larger positive and negative acceleration levels. Gap search component : In general, drivers do not begin searching for a gap in the ma inline traffic until they can see the nose of the entrance ramp ( 10 ). On cloverleaf ramps, drivers focus on navigating the curves until they are on a transition spiral to the straight portion of the lane ( 10 ). Merge steering component: Two types of me rges have been described in th e literature. The first, a short me rge, occurs when the driver me rges before or near the end of the entran ce ra mp nose. This type of merge is likely to occur on ra mp s with poor geometry where drivers merge aggressively to avoid being trapped at the end of the lane ( 11 ). The second type, the long me rge, occurs when the driver uses almost the enti re length of the acceleration lane. This type of me rge occurs when the geometry is good and traffic volumes are high ( 11 ). Sarvi et al. ( 6 ) suggest that when there is heavy congestion, gap search does not actually occur; instead, zip me rging happens at the end of the ra mp (i.e., where ramp and freeway vehicles merge one by one in an alternating pattern). There is some disagreement as to which geometric component has the greatest effect on gap acceptance: right lane volumes ( 11 ), ramp design ( 6 ), or gap distribution ( 4 ). Overall, Hunter and Machemehl ( 7 ) found that 85% of entering vehicles merged comfortably in 375 m as measured fro m the point where the ra mp and mainline pavement edges are 1.25 m apart to the end of the taper. Limited-length SCLs over 425 m are not necessary to improve me rging behavior. Abort component: The abort maneuver only occurs if drivers do not find a suitable gap within the length of the SCL. Their focus changes from gap search to an avoidance ma neuver and they decelerate to stop before the end of the SCL ( 4 ). De si gn Is su es Merging speeds of elderly drivers are lower than those of younger drivers when there are no cars in the ma inline right lane. These speeds decreased further when there were cars on the through lane. Merge point distributions were similar for young and old drivers ( 12 ). Cr os s Re fe re nc es Key Co mp onents of Sight Distance, 5-2 Ke y Re fe re nc es 1. McCartt, A.T., Northrup, V.S., & Retting, R.A. (2004). Types and characteristics of ra mp -related motor vehicle crashes on ur ban interstate roadways in northern Virginia. Journal of Safety Research, 35 (1), 107-114. 2. De Waard, D., Dijksterhuis, C., & Brookhuis, K.A. (2009). Merging into heavy motorway traffic by young and elderly drivers. Accident Analysis & Prevention, 41 (3), 588-597. 3. Ahammed, A.M., Hassan, Y., & Sayed, T.A. (2006). Effect of geometry of entrance te rm inals on freeway merging behavior. Transportation Research Board 85 th Annual Meeting Co mpendium of Papers [CD-ROM]. 4. Michaels, R.M., & Fazio, J. (1989). Driver behavior model of merging. Transportation Research Record, 121 3 , 4-10. 5. Choudhury, C.F., Ramanujam, V., and Ben-Akiva, M.E., (20 09). Modeling Acceleration Decisions for Freeway Merges, Transportation Research Board 88th Annual Meeting Compendium of Papers [CD-ROM]. 6. Sarvi, M., Ceder, A., & Kuwahara, M. (2002). Modeling of freeway ram p me rging process observed during traffic congestion. Proceedings of the 15 th International Symposium on Transportation and Traffic Theory, 483-502. 7. Hunter, M., & Machemehl, R. (1999). Reevaluation of Ramp Design Speed Criteria: Review of Practice and Data Collection Plan (FHWA/TX-98/17321-1). Austin, TX: Center for Transportation Research. 8. Fukutom e, I. and Moskowitz, K., (1960). Traffic Behavior and On-Ramp Design, Highway Research Board Bulletin #235 . 9. Kondyli, A., and Elefteriadou, L. (2009). Driver Behavior at Freeway-Ramp Merging Areas: Focus Group Findings, Transportation Research Board 88 th Annual Meeting Compendium of Papers [CD-ROM]. 10. Bhise, V.D. (1973). Visual search by drivers in freeway me rgin g: Im plications for vehicle design. Proceedings of the Human Factors Society Annual Meeting, 17 (1), 152-161. 11. H unter, M., Machemehl, R., & Tsyganov, A. (2001). Operational evaluation of freeway ram p design. Transportation Research Record, 1751 , 90-100. 12. Kimura, K., & Shimizu, K. (1998). Driving behavior of the elderly at merging sections. Proceedings of the 8 th International Conference on Transport and Mobility for Elderly and Disabled People, 109-117. 12-3 HFG INTERCHANGES Version 2.0
R ED UC IN G W RO NG -W AY E NT RI ES ON TO F RE EW AY E XI T R AM PS In tr od uc ti on Reducing wrong-way entries onto freeway exit ramps refers to treatments that can be used to reduce the frequency of dr iv ers entering freeways by using the exit ramps. An average of 350 fatalities occur each year in the United States as a result of wrong-way crashes on freeways ( 1 ). Furtherm ore, exit ram ps were found to be the most frequent origin of wrong-way incidents. In the sam ple of wrong-way dri vers, elderly drivers are overrepresented by experiencing twice the wrong-way crashes than would be expected. Crashes often occur in the early morning hours, although this may be linked with the high frequency of impaired drivers. De si gn Gu id e lin es This guideline can be used to identify roadway treatments or geometric countermeasures that address specific issues contributing to wrong-way driving, thereby reducing the occurrence of drivers entering freeways via exit ra mp s. Issues Contributing to Wrong-Way Driving (2) Roadway Treatments or Geometric Countermeasures Light land use ⢠Increase conspicuity, e.g., use wrong-way arrows or red reflectorized raised pavem ent markings (RRPMs) ( 3 ) ⢠Monitor interchanges in areas of light land use and low traffic volumes Low traffic volum es Poor visibility ⢠Increase/improve roadway lighting Adequate directional signing (except at driveways) ⢠Lower âDo Not Enterâ and âWrong Wayâ signs ( 4 ) Confusing or poorly visible access point configurations ⢠Avoid freeway left-side exit ramps ( 4 ) ⢠Provide mo re cues using the ra mp geom etry (e.g., mo re severe angles on the right side as the vehicle passes the exit ramp on the roadway; 5 ) ⢠Median installations ( 6 ) G EO ME TR IE S TO D IS CO UR AG E W RO NG -W AY E NT RY Di vi de d Cr os sr oa ds Tw o- La ne Cr o ssr oa ds Source: recreated fro m AASHTO ( 5 ) Based Primarily on Expert Ju d g ment Based Equally on Exper t Judgment and Empirical Dat a Based Primarily on Empirical Dat a HFG INTERCHANGES Version 2.0 12-4
Di scu ssi on All of the characteristics listed in the guideline on the previ ous page correspond to mi ssing cues that can help drivers deter mi ne that they have start ed down the roadway, particularly an exit ramp, in the wrong direction. Light land use and low traffic volumes : Wrong-way driving crashes tend to occur in areas with light land use and low levels of traffic. Both of these situations likely indicate occasions when few car s would be traveling in the opposing direction to signal to drivers that they are going to be traveling the wrong way. In these cases, drivers need another indication that they have started traveling in the wrong direction. In a laboratory study of straight mu ltilane roads, Miles, Carlson, Ullm an, and Trout ( 3 ) found that the addition of directional arrows or red RRPMs led to more correct identifications of the proper travel direction on a roadway, though the effects were moderate. Poor visibility : Wrong-way movements tend to occur when visibility is poor ( 2 ). In a study of wrong-way crashes, 74% were found to occur during the dark hours of the day. An obvious potential solution is to increase the level of artificial lighting at the access points. An increase in lighting levels would make some of the cues that are available to drivers more apparent at nighttime. Adequate directional signing: Scifres and Loutzenheiser ( 2 ) found that in most cases, the signing at most of the origins was adequate (with the exception of driveways). However, signing improvements have been suggested by Cooner, Cothron, and Ranft ( 4 ). Notably, they refer to a lower mounting of âDo Not Enterâ and âWrong Wayâ signs, shown to be effective in the state of California. The lower height avoids sight restrictions, is in the range of low-beam headlights for night driving, and is potentially more visible to impaired drivers who drive with their eyes on the pave me nt. The bottom of the sign package is installed 2 ft above the edge of the pavem ent (though this is inconsistent with the MUTCD). Confusing or poorly visible access point configurations : Wrong-way m ove me nts were increased when the design of the access point was difficult to see or understand ( 2 ). The si mp licity of the access points can be im proved in mu ltiple ways. The first is to avoid the construction of left-side exit ram ps and install reflectorized wrong-way pavem ent arrows on existing left-side ramps ( 4 ). During a crash analysis, left-side exits experienced multiple crashes due to wrong-way en tries. Driv ers are familiar with turning right to enter a freeway and may end up traveling the wrong way up a left exit ram p by using th is maneuver. Additionally, AASHTO ( 5 ) recommends sharp or angular intersections between the crossroad and the ramp, making the incorrect maneuver less natural to execute. Median islands can also make incorrect turning movements more difficult. The diagrams on page 12-4 provide sample road geom etrics that discourage wrong-way entry. De si gn Is su es Drivers who are under the influence of alcohol and/or other drugs com pose a considerable portion of wrong-way dr iv ers. Although these drivers cannot specifically be designed for, countermeasures that reduce the affordance of driving the wrong way (such as geom etric alterations) may be mo re effective than those which require the perceptual abilities of the drivers to function at a certain level (such as signage or pavement markings). These trade-offs can be considered in areas near bars or other locations where drunk drivers may be mo re prevalent. Short sight distance has been found to be a contributor to wrong-way crashes ( 7 ). Improving sight distance may decrease the nu mb er of drivers driving the wrong way by increasing the odds that they will see an approaching right- way driverâs headlam ps. However, im proving sight distance is mo re of a crash avoidance m easure for right-way dr iv ers who will see wrong-way drivers approaching from a greater distance. Cr os s Re fe re nc es Lighting Guidelines, 21-1 Ke y Re fe re nc es 1. Cooner, S.A., & Ranft, S.E. (2008). Wrong-way driving on freeways: Problems, issues, and countermeasures. Transportation Research Board 87th Annual Meeting Compendium of Papers [CD-ROM]. 2. Scifres, P.N., & Loutzenheiser, R.C. (1975). Wr ong-W ay Movements on Divided Highways. (JHRP-13-75). West Lafayette, IN: Purdue University . 3. Miles, J.D., Carlson, P.J., Ullman, B.R., & Trout, N.D. (2008). Red retroreflective raised pavement markings: Driver underst anding of their purpose. Transportation Research Record, 205 6 , 34-42. 4. Cooner, S.A., Cothron, A.S., & Ranft, S.E. (2004). Countermeasures for Wrong-Way Movement on Freeways: Guidelines and Recommended Practices. (FHWA/TX-04/4128-2). College Station: Texas Transportation Institute. 5. AASHTO (2011). A Policy on Geometric Design of Highways and Streets. Washington, DC . 6. Moler, S. (2002). Stop. You're going the wrong way! Public Roads, 66 (2), 24-29. 7. Rinde, E.A. (1978). Off-Ramp Surveillance (FHWA-RD-79-T0334; 623124). Sacramento: California Depart me nt of Transportation. 12-5 HFG INTERCHANGES Version 2.0
DRIVER EXPECTATIONS AT FREEWAY LANE DROPS AND LANE REDUCTIONS Introduction Matching driver expectations at freeway lane drops is important because lane drops represent a situation that may violate driver expectations and cause confusion when th e driver expects the lane to continue on the freeway mainline. This confusion can result in high speed variability, erratic maneuvers, and driver frustration ( 1 ). Additionally, a left lane drop situation violates multiple driver expectations and can cause more problems. All of these results have negative safety implications; thus, th e more accurately that lane drops conform to driver expectations, the safer the situation will be. This guideline re fers specifically to lane drop s on freeway sections that do not include exit ramps. Design Guidelines Consideration should be given to the following principles related to lane drops and lane reductions to promote driver behavior that is consistent with the safe use of such facilities (based on Goodwin ( 2 )). Principle Guideline Visual Principles Provide continuous visibility The minimum distance that should be visible to a driver is that required to: a. Perceive that the lane is ending, b. Evaluate maneuver options, and c. Maneuver to an adjacent lane. Minimize attention-dividing conditions Place the lane drop away from other distracti ons such as ramps or complicated signage. Provide adequate transition cues Provide a taper of sufficient length so that drivers who enter it with no knowledge of the lane drop will have suffici ent time to maneuver. Coordinate the visual and operational drop Create a lane reduction such that the lane does not appear to continue beyond the operational reduction, even if the pavement does continue. Geometric Principles Provide an adequate escape area Provide an adequately sized escape area at ex it lane drops for driver s who have insufficient time before the exit gore to make a normal lane change. Signing Principles Notify the driver that the lane is not continuous Warn drivers who enter the freeway by using an add-drop lane that the lane is not a permanent addition. Use adequate traffic control devices Use adequate and consistent traffic control devices to inform drivers: a. What is going to happen b. Where it is going to happen, and c. What they need to do. S PECIFIC G UIDANCE FOR L ANE D ROPS AT E XITS WITH O PTION L ANES For lane drops with option lanes, clearly communicate ( 3 ): 1. The dropped lane can only reach the exit 2. The option lane leads to either the exit destination or the mainline 3. Any other lane only reaches the mainline 4. The identifying information fo r each destination (e.g., stre et name, destination name) Based Primarily on Expert Jud g ment Based Equally on Expert Judgment and Empirical Data Based Primarily on Empirical Data HFG INTERCHANGES Version 2.0 12-6
Discussion It should be noted that the lane drop guidance provided by Goodwin (2) is not specific to exit lane drops as they are commonly referred to today. This guidance was also formulated based on the study of lane reductions on freeways. Visual principles: Lane drops should be located where drivers can see them continuously for a long enough period of time to perceive that the lane is about to end, decide on a maneuver, and execute that maneuver. Thus, lane drops should not be located just over the crest of a vertical curve or around a horizontal curve (2). Lane drops should be located away from other conditions which require the driverâs attention, as these locations increase the probability of drivers missing lane drop cues while looking at other roadway features. A major visual cue for navigating lane drops is the lane drop taper. An inappropriately short taper may cause drastic lane changes, while an overly long taper does not provide cues that the lane is ending. From the driverâs viewpoint, operational and physical lane reductions should be coordinated. So, if the pavement continues beyond the operational lane drop, it should be apparent that the lane does not continue onto that pavement. Geometric principles: Cornette (4) found that lane drops, lane splits, and lane reductions at sites with poor geometrics (i.e., high rates of curvature, sight distance restrictions) had higher conflict rates than those at sites with better geometric features. Drivers who do not expect the lane drop should have a reasonable opportunity to recover and stay on their route. An adequate escape area should be provided at/after an exit lane drop to provide drivers who do not want to exit a chance to recover and remain on the mainline (2). Signing principles: A subset of the drivers who encounter a lane drop may have just entered the freeway. If drivers are able to enter using an add-drop lane, they should be warned that their lane is not continuous for through travel (2). For all drivers, adequate and consistent information should be provided by the traffic control devices. It is important for drivers to know if they are required to take an action, or if other drivers are required to act. Additionally, excess information not related to the lane drop segment should be minimized (2). Lane drop exits with option lanes provide particularly difficult circumstances for drivers. It is often unclear to drivers that the option lane serves both the mainline and the exit destinations. The underutilization of the option lane can lead to a loss of service volume for the roadway as well as a number of unnecessary merge maneuvers. Design Issues On United States border roadways that are used by a large percentage of Spanish-speaking drivers, additional signage with Spanish legends may be appropriate. To convey the message âRight Lane Ends,â the sign that had the highest overall comprehension level among Spanish-speaking drivers was âCarril Derecho Terminaâ (5). This sign also had the highest comprehension levels among English-speaking drivers among the three Spanish-legend signs that were tested. Dynamic late-merge systems have been developed for use in work zone lane closure situations. These systems utilize a series of changeable message signs and static work zone signs to provide merge information to the driver based upon the current traffic volume through the work zone. The basic principle supports early merging when the traffic flow is light and late merging (closer to the gore point) when the traffic volume is heavier. Cross References Passing Lanes, 16-2 Effectiveness of Symbolic Markings, 20-4 Key References 1. Chrysler, S.T., Williams, A.A., Funkhouser, D.S., Holick, A.J., & Brewer, M.A. (2007). Driver Comprehension of Diagrammatic Freeway Guide Signs. College Station: Texas Transportation Institute. 2. Goodwin, D.N. (1975). Operational effects of geometric design at freeway lane drops (Abridgment). Transportation Research Record, 541, 26-30. 3. Upchurch, J., Fisher, D.L., & Waraich, B. (2005). Guide signing for two-lane exits with an option lane. Transportation Research Record, 1918, 35-45. 4. Cornette, D.L. (1972). Operational Characteristics of Lane Drops. (KYHPR-70-63, HPR-1(18), Part II). Lexington: Kentucky Bureau of Highways. 5. Hawkins, H.G., Jr., Picha, D.L., Kreis, D.C., & Knodler, M.A. (1999). Evaluation of Alternative Traffic Signs for Use in Texas Border Areas. (FHWA/TX-99/1274-3). College Station: Texas Transportation Institute. 12-7 HFG INTERCHANGES Version 2.0
DRIVER INFORMATION NEEDS AT COMPLEX INTERCHANGES Introduction Accommodating driver expectations at interchanges is paramount to navigational success. Expectations refer to âa driverâs readiness to respond to situations, events, and information i n predictable and successful waysâ ( 1 ). Complex interchanges should be designed to give the drivers what they expect to see ( 2 ). Information that reinforces expectancies helps drivers respond faster, whereas information that violates expectancies leads to longer task times and/or errors ( 3 ). Thus, more predictable design and operation leads to fewer errors ( 4 ). Design Guidelines Geometric Elements Route continuity: ⢠Provide a route on which changing lanes is not necessary to continue on the through route ( 5 ) . ⢠If possible, provide the greatest number of lanes for the through movement ( 6 ) . Lane balance: ⢠âThe number of lanes leaving a diverge point is equal to the number of lanes approaching it, plus oneâ ( 7 ; see figure ). ⢠Minimize the required number of lane shifts by using option and auxiliary lanes ( 6 ). Ramp spacing: ⢠Provide adequate ramp spacing to allow for clear and simple guide signing, and to prevent congestion from heavy traffic entering and exiting ( 6 ). Error handling: ⢠Provide a forgiving roads ide at critical features ( 2 ). ⢠Avoid creating compound geometric features ( 2 ). Signing Error handling: ⢠Eliminate information - related error sources: avoid deficient, ambiguous, confusing, missing, misplaced, blocked, obscured, small, illegible, or inconspicuous displays ( 3 ) . Sign placement: ⢠Spread out competing information sources by moving less important information upstream or downstream ( 3 ). ⢠Structure driver expectations through advanced warning ( 4 ) . ⢠Repeat important information or do not provide interchange information so far upstream that it is forgotten by the time that the interchange is reached ( 3 ). Sign content: ⢠Provide appropriate signing to guide drivers ( 2 ) . ⢠Satisfy all driver information needs ( 3 ) . Sight Distance ⢠Avoid sightline restrictions ( 1 ) . ⢠Provide visibility that is proportional to feature criticality ( 2 ) . M AJOR W EAVE WITH L ANE B ALANCE AT E XIT G ORE Source: Transportation Research Board ( 7 ) Based Primarily on Expert Judgment Based Equally on Expert Judgment and Empirical Data Based Primarily on Empirical Data Number of lanes leaving the diverge point equals the number of lanes approaching it, plus one. HFG INTERCHANGES Version 2.0 12-8
Di scu ssi on Overall, there is little information available in the research literature that provides specific guidance related to supporting driver expectations at interchanges. The information provided in the guideline on the previous page is generally related to principles of geometry, signage, and sight distance to s upport elements of driver expectations at interchanges. Geometric elements: Doctor, Merritt, and Moler ( 6 ) discuss driver expectations in reference to mu ltiple ele ments of interchange design. Ramp spacing that is too close can lead to congestion and cluttered signage. The combination of system and service interchange s can lead to info rm ation overload, inconsistent sign design, and contradictory movements. Route continuity is provided by designing a roadway on which âchanging lanes is not necessary to continue on the through routeâ ( 5 ). This principle âreduces lane changes, simplifies signing, delineates the through route, and reduces the driverâs search for directional signingâ ( 5 ). Additionally, drivers sometim es assu me th at at a split, the leg with the greater nu mb er of lanes carries the ma in route. To use lane balance, desig ne rs arrange the lanes on the freeway to require drivers to take the mi nimu m num ber of lane shifts. This is done by using option and auxiliary lanes. Signing: Advance guide signing that prepares drivers to ma ke decisions and maneuvers is possibly the most im portant strategy for helping drivers navigate complex interchanges ( 6 ). Signing can spread out the amount of ti me that drivers can use to perform lane changes in advance of the decision points. Additionally, signs at the decision point confirm decisions that drivers made on the approach ( 6 ). Lunenfeld ( 3 ) stresses that the amount of inform ation should not overwhelm the driver. W hen inform ation sources compete, the less important sources should be mo ved upstream or downstream . However, information should not be provided so far upstrea m th at it is forgotten by the tim e that the interchange is reached. If inform ation is too far upstream, it may need to be repeated closer to the interchange. Sight distance: Adequate sight distance is required due to the reliance on visual information and for co mp lex decision ma king ( 3 ). Sightline obstructions that cover up im portant or critical inform ation cues should be avoided. De si gn Is su es In particular, driver expectations can be easily violated at transition sections where the roadway conditions change. Drivers anticipate the upcoming roadway characteristics based on features that are common to the road they are on ( 4 ). Roadway designers should look for possible expectancy violations where changes in roadway characteristics (e.g., geometrics, design, or operation) or changes in operating practices o ccur (e.g., speed zones, no passing zones, or signal timings; 1 ). Features that are âfirst of a kindâ on a particular roadwa y or t hose that drivers may find unusual or special are also im portant to exa mi ne ( 1 ). Additionally, adequate transitions should be provided ( 2 ). Cr os s Re fe re nc es Sight Distance Guidelines, 5-1 Driver Expectations at Freeway Lane Drops and Reductions, 12-6 Signing Guidelines, 18-1 Ke y Re fe re nc es 1. Lunenfeld, H., & Alexander, G.J. (1 984). Hu ma n factors in highway design and operations. Journal of Transportation Engineering, 110 (2), 149-158. 2. Messer, C.J., Mo unce, J.M., & Br ackett, R.Q. (1981). Highway geometric design consistency related to driver expectancy: Vol. I, Executive summary. (FHWA/RD-81/035). Wa shingto n, DC: FHWA. 3. Lunenfeld, H. (1993). Hu ma n factors associated with interchange design features. Transportation Research Record, 1385 , 84-89. 4. Cline, E.L. (1999). Adherence to design standards and guidelines: The hum an factor. 69 th ITE Annual Meeting. 5. AASHTO (2011). A Policy on Geometric Design of Highways and Streets . Washington, DC. 6. Doctor, M., Merritt, G., & Moler, S. (2009). Designing co mp lex interchanges. Public Roads, 73 (3), 3-11. 7. Transportation Research Board (2000). Highway Capacity Manual 200 0 . Washington, DC. 12-9 HFG INTERCHANGES Version 2.0
A RRO W - PE R -L AN E S IG N D ES IG N TO S U PPO RT D RI VE R N AV IG AT IO N In tr od uc ti on Arrow-per-lane (APL) signs are com posed of prim arily two parts: arrows that point to individual lanes, and the destination inform ation listed above the arrows. They gene rally either are large signs or occur in groups, because every individual lane requires its own arrow. These signs provide clear direction for the destinations reached by each lane and have been shown to increase the num ber of correct lane choices by ol der drivers as compared to standard diagrammatic signs ( 1 ). The signs shown in the guideline vary slightly from those recommended by the MUTCD ( 2 ), because they are m odeled after real sign exam ples. Therefore, this guideline is focused on troubleshooting driver issues rather than informing new sign design. De si gn Gu id e lin es To effectively support driver navigation, the destination in formation and sign design must allow drivers to pair the destination inform ation with an arrow, and consequently, an arrow with a travel lane. P AI RI NG D ES TI NA TI ON I NF OR MA TI ON WI TH O NE OR M OR E A RRO WS Poorly Distinguished Information Easily Associated Information Unsymmetrical centered text above split text: Text centered above one or more arrows: ⢠30 We st shield can be interpreted to apply only to left lane. ⢠Centered text is easily ma tched to one or mo re arrows. Hyphenated destinations: Stacked and centered text: ⢠Hyphenated destinations ma y cause driver confusion. ⢠Stacked destinations are interpreted to go with both arrows. âExit Onlyâ by one of multiple arrows : Exit placard centered above a panel: ⢠âExit Onl yâ notation may be associated with the destination rather than the arrow. ⢠A centered exit placard is interpreted to apply to the entire sign. P AI RI NG A RR OW S WI TH T RA VE L L AN ES Causes of Driver Confusion What to Do to Fix It Arrows do not appear to be centered over the lanes. On tangents, make sure that the arrows are centered over the lanes fro m the ti me when the sign is first legible until the driver passes the sign (for legibility distance calculations, see Tutorial 5). Avoid APL signs on sharp horizontal curves. All of the destinations above an arrow are not reachable by using that lane. Avoid positioning a destination above an arrow if it canât be reached by the indicated lane. All of the destinations above an arrow are not able to be reached by following the same direction at a split or option lane. Match the layout of the destination inform ation to the roadway geometry. Based Primarily on Expert Judgmen t Based Equally on Expert Judgment and Empirical Dat a Based Primarily on Empirical Da ta HFG INTERCHANGES Version 2.0 12-10
Di scu ssi on The driversâ reading goal is to associate destination information with an arrow which points to a particular lane. To accom plish this goal, the pertinent destination inform ation mu st be able to be separated from adjacent information and clearly pertain to one or mo re arrows. The prim ary sign feature that accom plishes this is the centering of information above the applicable arrows . In essence, destinations are interpreted as being centered above the arrow(s) to which they apply. Design elements that cause inappropriate continuity or separation between elements make the navigation task more difficult. The figure on page 12-10 shows both good and bad examples of pairing destination inform ation with arrows. Poorly distinguished informatio n : Sign layout can make destination information more difficult to associate with the relevant arrows. One way in which this may occur is when destination information that is meant to be shared by mu ltiple arrows (i.e., mo re than one lane leads to the same destination or destinations) is interpreted by drivers to mean that each lane leads to a different destination. For example, with the 30 West notation in the guideline, the entire text segment is centered on the sign. Some drivers, however, may interpret the route shield as being more toward the left side of the sign and associate the route with only the left-hand arrow rather than bo th arrows because the route shield is much larger than the âwestâ notatio n. Another problem ma y occur when text and associated arrows that are meant to indicate different destinations are interpreted by drivers to indicate multiple lanes that lead to the same destination. Destination groupings that are separated by a hyphen may lead to this kind of confusion. The destinations can be interpreted as two areas on the sa me roadway, or two separate areas. The hyphen prevents association with a single arrow by creating continuity be tween destin ation nam es. Thus, it should be avoided. In Richard and Lichty ( 3 ), a situation arose where the âExit Onlyâ indication was positioned above a single lane on a mu ltiple-APL sign. Drivers interpreted this âExit Onlyâ text in a different way than the destination information when used above a single arrow; rather than applying this information to the arrow, and thus the travel lane, drivers assigned it to the destination(s). Therefore, this positioning may cause some drivers to believe that they need to be in the exit-only lane to reach the destination, causing unnecessary lane changes. Easily associated information : In Richard and Lichty ( 3 ), destination information shown directly above an arrow was associated by all drivers with the lane below the arrow. Drivers thought that an arrow with multiple destinations m eant that all of the listed destinations could be reached using that lane. Similarly, information that was centered above multiple arrows was generally seen as applying to all of the arrows that it was centered above. This applied to destination inform ati on above mu ltiple arrows or sm aller pl acards above larger guide signs that referred to the entire guide sign. The signs included in the guideline section on the previous page are m odeled after real sign exam ples. It should be noted that they vary slightly from those recommended in the MUTCD ( 2 ). For exam ple, those included show downward pointing arrows, whereas the MU TCD recommends upward poi nting arrows except in cases where the lane use is restricted to the listed destinations. The direction that the arrows point should not have an impact on the visual groupi ng performed by drivers. The MUTCD has additional guidance recommending using only one destination per m ove men t; however, multiple destinations listed per movement are common on existing signs. This gui deline does not seek to contradict the MUTCD with design guidance, but rather to discuss general interpretation patterns and provide guidance for troubleshooting problem atic signs and prioritizing their replacem ent. De si gn Is su es Separations between sign panels can also be used to distinguish between lanes or lane groupings. Destinations on different panels are typically not associated with one another ( 3 ). Splits between sign panels also show where âExitâ placards apply and do not apply. The MUTCD ( 2 ) describes the usage of a vertical white line to separate diverging route m ove me nts. This is likely co mp arable to the distinction between separate sign panels. Cr os s Re fe re nc es General Principles for Sign Legends, 18-2 Driver Co mp rehension of Signs, 18-8 Ke y Re fe re nc es 1. Golem biewski, G., & Katz, B.J. (2008). Diagrammatic Freeway Guide Sign Design. Traffic Control Devices Pooled Fund Study. Retrieved July 2011 fro m http://www.pooledfund.org/projectdetails.asp ?i d=281&status=4. 2. FHWA (2009). Manual on Uniform Traffic Control Devices for Streets and Highways . Wash ington, DC. 3. Richard, C. and Lichty, M.G. (2011). Driver expectations when navigating co mp lex interchanges. Task 4: Gather feedback from dr ivers. Draft report (Federal Highway Ad mi ni stration Contract No. DTFH61-08-D-00032-T-10005) . Seattle, WA: Battelle. 12-11 HFG INTERCHANGES Version 2.0
DRIVER BEHAVIORAL TRENDS BASED ON EXIT RAMP GEOMETRY Introduction Exit ramps provide the means of accessing adjacent surface streets from a freeway. Well-designed exit ramps provide sufficient area for vehicles to depart from the main freeway lanes and sufficient distance for vehicles to decelerate comfortably from freeway speeds to a speed appropriate for the controlling feature of the ramp, which may be the first curve encountered along the ramp or it could be the crossroad terminal. Driver behavior at freeway exit ramps is based upon a variety of factors, including the operating conditions along the freeway and the geometry of the ramp. Design Guidelines To design exit ramps, it is important first to define the intended behaviors of an exiting driver (1). Ramps should be designed accordingly to support these safe driving behaviors. The figure highlights where key driver behaviors/decisions take place in the vicinity of an exit ramp. The numbers in the figure correspond to the driver behaviors listed in the first column of the table. Driver Behaviors for Safe Exit Design Features to Support Safe Driving Behaviors 1. The driver should maintain a relatively constant speed in the freeway lanes. 2. The driver should position his/her vehicle in the right lane of the freeway prior to the beginning of the deceleration lane. 3. The driver should signal to indicate his/her intended maneuver to other drivers in the traffic stream. 4. The driver should initiate the diverge maneuver shortly after the deceleration lane begins. 5. Deceleration should begin gradually, immediately after entering the deceleration lane. 6. The driver should reach ramp speed before the end of the deceleration lane. 1. Proper sequence and location of guide signs to allow drivers time to make proper route choice decisions. 2. Sufficient sight distance to allow drivers to perform appropriate maneuvers. 3. Pavement markings and roadside delineation to delineate the proper trajectory along the ramp. 4. Delineation to distinguish the features of the gore area. 5. For taper-type exits, sufficient divergence angle to provide a clear indication of the point of departure from the through lanes. A typical divergence angle is usually between 2 and 5 degrees. 6. For parallel-type exits, a taper area should be provided to indicate the general path to be followed by the exiting driver. Typical taper lengths are between 15:1 to 25:1 [longitudinal:transverse]. 7. Deceleration lane lengths sufficient for drivers to reduce their speed from the operating speed along the freeway to the average running speed of the controlling feature at the end of the speed-change lane. Minimum deceleration lane lengths are provided in Table 10-5 of the Green Book (2). PLAN VIEW OF DRIVER BEHAVIORS AT EXIT RAMP* *Numbers correspond to âDriver Behaviorsâ in the guideline above. Based Primarily on Expert Judgment Based Equally on Expert Judgment and Empirical Data Based Primarily on Empirical Data HFG INTERCHANGES Version 2.0 12-12
De si gn Gu id an ce On an exit ramp, deceleration is accomplished first as the driver removes his/her foot from the throttle and the vehicle coasts in gear for a period of time (ty p ically ) without the use of brakes, and then the driver applies the brake s and de celerates comfortably . A recent study ( 3 ) confirmed that drivers coast in gear an average of 3 s prior to apply i ng the brakes to decelerate along a deceleration lane. Coasting time was defined to be the sum of the elapsed time between occurrence of peak speed and deactivation of throttle and t he elapsed time between deactivation of throttle and activation of br ake. Furtherm ore, drivers ty pically coast in gear approximate ly 2 s in the freeway lanes and approximately 1 s in the deceleration lane prior to applying the brakes. Through further investigations of diverge locations and speeds and deceleration, Torbic et al. ( 3 ) concluded the minimum deceleration lane lengths provided in the 2004 Green Book ( 4 ) are conservative estimates, given the current vehicle fleet and driver population. Drivers decelerate at levels well within the capabilitie s of the vehicle fleet and driver preferences. This is, in part, due to some deceleration by drivers in the freeway prior to the diverge maneuver. Drivers typically diverge between 4 to 7 mi/h below average freeway speeds; however, it is prudent for designers to assume that all deceleration takes place in the speed-change la ne when determining minimum deceleration lane length. Although it seems intuitive that a relationship should exist between deceleration level and deceleration lane length, no relationship has been determined ( 5 ). It has been found that longer deceleration lanes lead to later deceleration at a higher level, perhaps because drivers relax thinking there is more time than there actuall y is to decelerate ( 5 ). Also, as deceleration lane length increases, the percentage of return maneuvers increases ( 6 ). On the other hand, shorter deceleration lanes lead to an increase in early exits ( 6 ) and deceleration along the taper to the lane. Torbic et al. ( 3 ) found similar results, indicating that providing deceleration lanes longer than the minimum values provided in the 2004 Green Book ( 4 ) ma y promote more casual deceleration by exiting dri v ers, particularly under uncongested or lightly congested conditions, but noted thi s is not necessarily a negative result. Simply, it changes the operational characteristics of the ramp. Most drivers diverge from the freeway either within the taper or the first two-thirds of the speed-change lane (defined as the distance between the end of the taper to the painted nose). Few drivers diverge from the freeway in the final third of the speed-change lane or bey ond the painted nose. Drivers that diverge earlier along the speed-change lane decelerate at a more casual level compared to drivers that diverge closer to the painted nose ( 3 ). De si gn Is su es The current design criteria for exit ramps assume free-flow or uncongested conditions along the freeway and are based upon the vehicle capabilities of passenger cars and driver comfort levels. Several studies ( 7, 8 ) recommend longer deceleration lane lengths on the order of 15% to 50% longer than those required for passenger cars to better accommodate the reduced vehicle capabilities of heavy vehicles. However, when exiting the freeway, trucks deceler ate at levels very comparable to those of passenger cars ( 3 ). In addition, truck drivers typically choose to diverge from the freeway at lower speeds than drivers in passenger cars and, during uncongested freewa y conditions, the distribution of diverge locations for trucks is very similar to the distribution of diverge locations for passenger cars. One of the goals in designing an exit ramp should be to minimize erratic behaviors near the ramp such as crossing gore paint, crossing gore area, stopping in gore, backing up, sudden slowing, lane changing (to exit), swerving, and stopping on shoulder. Erratic maneuvers occur mo st frequently after lunch, after rush hour, and during the first hour of darkness during mid-morning and mid- afternoon. These data suggest that most erratic maneuvers are made by motorists taking unfamiliar routes, as contrasted to shop ping or commuting-to-work trips, which involve familiar and frequently used r outes ( 9 ). Proper sequencing and location of overhead guide signs, good delineation of the exit ramp, and clearl y distinguishing the taper, the beginning of the deceleration lane, a nd the gore area using pavement markings (e.g., raised pavement markers) and roadside delineation, in addition to geometrics, need to be considered to reduce driver confusion near exit ramps. Cr os s Re fe re nc es Task Analysis of Driver Merging Behavior at Freeway Entrance Ramps, 12-2 Ke y Re fe re nc es 1. Mace, D.J., Hostetter, R.S., & Sequin, L.E.. (1969). Information Requirements for Exiting at Interchanges . (Report 89211-f). State College, PA: HRB-Singer. 2. AASHTO (2011). A Policy on Geometric Design of Highways and Streets . Washington, DC. 3. Torbic, D.J., Hutton, J.M., Bokenkroger, C.D., Harwood, D.W., Gilm ore, D.K., Knosha ug, M.M., et al. Design Guidance for Fre e way Mainline Ram p Ter mi nals . F inal Report. NCHRP Project 15-31A. Washington, DC: Midwest Research Institute. 4. AASHTO (2004). A Policy on Geometric Design of Highways and Streets . Washington, DC. 5. Bella, F., Garcia, A., Solves, F., & Ro me ro, M.A. (2007). Driving si mu lator validation for decele ration lane design. Transportation Research Board 86th Annual Meeting Compendium of Papers [CD-ROM ] . 6. Garcia, A., & Rom ero, M.A. (2006) Experim ental obser va tion of vehicle evolution on dece lera tion lanes with different length s . Tr ansportation Research Board 85th Annual Meetin g Compendium of Papers [CD-ROM] . 7. Ervin, R., Barnes, M., MacAdam , C., & Sc ott, R. (1986). Im pac t of Specific Geometric Features on Truck Operations and Safety at Interchange s (FHWA/RD- 86/057). Wa shington , DC: FHWA. 8. Firestine, M., McGee, H., & Toeg, P. (1989). Improving Truck Safety at Interchange s (FHWA-IP-89-024). Washington, DC: FHWA. 9. Taylor, J.I., & McGee. H.W. (1973). NCHRP Report 145: Improving Traffic Operations and Safety at Exit Gore Areas . Washington, DC: Highway Research Board. 12-13 HFG INTERCHANGES Version 2.0
