Guidebook for Advancing Collaborative Decision Making (CDM) at Airports (2015)

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.

32 After describing the background and general processes for implementing a project at an air- port using the ACDM process, it is important to acknowledge and discuss the myriad challenges that may be encountered in such a project. Some of these challenges and limitations have been encountered in the FAA/Industry CDM program; others have been encountered in actual field trials, while others are a result of the ACDM activity in Europe. These challenges and differences will vary from airport to airport. Lessons Learned There have been some significant examples where collaboration was not fully accomplished: • In May 2009, flight departures at a major airport in the northeast were restricted due to weather. This restriction was not communicated, and more than 70 aircraft taxied out, resulting in air- craft holding in departure queues on taxiways for hours. • During the winter of 2011, a snowstorm at a major airport in the northeast caused multiple flight diversions to another regional airport. The regional airport was unable to accom- modate the unexpected flight diversions. Passengers were held in aircraft on taxiways for hours. • In 2011, arrival flights at a major airport on the East Coast could not access an offloading gate because all ramp space was occupied by weather-delayed departure flights. • During the summer of 2013, an airport in the southwest received unexpected flight diversions due to severe weather at their original destination. These diversions were generated by two different airlines. The airport did not have the facilities to promptly refuel these flights, caus- ing lengthy delays. The first obvious lesson is the difference in airports that ACDM must accommodate. Every airport is different in type of operations, runway layout, seasonal weather types, unique ter- minal and parking position layout, density of operations, type and number of ramp tower operations, and multiple other differences. These local differences negate the effort for one standard ACDM operation among airports. Every airport has its individual needs, and these needs must be reflected in that airport’s ACDM program. Both U.S. ACDM and European ACDM efforts have encountered these differences. For example, each significant flight operator in the United States has some type of control center. These centers have a great deal of con- nectivity and collaborate in a centralized traffic management arena with the FAA ATCSCC and each other through the FAA/Industry CDM network. But they do not usually cooperate with each other on the airport level managing taxi queues and coordinating push-back. In Europe, the problem is the reverse, with the significant number of different ATC service providers C H A P T E R 5 Challenges and Limitations to Implementation

Challenges and Limitations to Implementation 33 usually airport-centric. There are some issues that are not obvious at the start but generally are encountered later: • Information sharing has to enable access to remotely located operations centers for airline, business aviation, and military operators. • Airline business case objectives are different. A hub and spoke carrier will have different needs than a point-to-point carrier. • Not all flights of the same carrier carry the same priority. The same airline might have differ- ent objectives for different flights for reasons that are unknown to every other stakeholder. • Open-mindedness is key. Unusual concepts that are not generally practiced can produce sig- nificant benefits. For example, airport and airline personnel working side-by-side in the same control facility can produce an effective workspace. Additionally, it can be expected that dif- ferent levels of participation might exist but usually are mitigated as the process matures and positive results experienced. IT Support/Integration Airports enabling ACDM concepts usually encounter the need to integrate information. Some- times historical local formats for systems such as FIDs seemingly dictate formats to be used. Stan- dard aeronautical data formats, such as AIXM and FIXM, should be utilized since flight operator operations centers will need to accept formats from multiple airports, and standard formats facilitate this process. Additionally, IT updates and the timing of such updates to systems used by more than one stakeholder require collaboration. Issues in this area are as follows: • Several individual airports have developed and installed “individual” airport systems of inte- grating flight data and ground tracking of aircraft. There is usually little coordination between airports to make these various systems uniform, so the flight operators are seeing the need to adapt to several different systems when they want real-time information from different airports. • The “individual” airport systems mentioned in the prior bullet are usually displayed on the computers of each individual flight operator. For optimum conformance in the airport sys- tem, each computer should be running the same version of software or known compatible versions. Thus the timing of the updating of these operator computers needs to be coordi- nated. This attribute may appear obvious but should not be underestimated. At an airport in the southern United States, construction data was being shared via PDF files. Some users updated their PDF reader and were able to read the files, while others did not and were not able to read the files. • Some airports simplify technology issues by using web pages. Most airports have restricted access web pages where ACDM information can be distributed. The use of this distribution method should be minimized wherever possible because operator operations center commu- nication might be inhibited by a large number of web pages to be accessed by several different airports. Trust between CDM Entities When the FAA/Industry CDM effort was established in the mid-1990s, information shar- ing was a new concept, and trust issues surfaced. The flight operators thought that universally exchanged delay data would be used by a competitor to illustrate their shortfalls in a public manner in order to gain market share. These trust issues resulted in some flight identification

34 Guidebook for Advancing Collaborative Decision Making (CDM) at Airports information being coded or scrambled so that a particular operator could only display the flight identifications of their operations. This complicated the quality assurance process because it required a central authority to unscramble all the data and interpret the results. Eventually, these trust issues were overcome once the stakeholders became aware of the benefits received from collaboration, and the data was unscrambled. Once the data was unscrambled, operators were held more accountable due to the fact that everyone was monitoring the same data, and thus collaboration actually increased. Positive collaboration leads to more collaboration. This was proven once again in the JFK airport departure metering initiative where originally conformance data was restricted to each individual operator. Once data indicating the degree of conformance was shared among all operators, conformance increased. It should be expected that when implementing ACDM, some trust issues will arise. This may be especially true where one large carrier is predominant at a particular airport. For example, this was true in the original early CDM and was overcome by the Ration by Schedule (RBS) concept, as detailed in the next section. Collaboration Barriers Collaboration barriers will always exist. Participants have different viewpoints and/or busi- ness objectives. For example, NBAA is a major participant in FAA/Industry CDM programs. But this association represents a number of different companies, some of whom do not desire, for competitive reasons, their company aircraft flight origin and destination information to be made public. A CDM workgroup developed the concept that individual aircraft owners could request that their data be masked. The data was included in all demand lists, but the actual aircraft iden- tification associated with that flight data was masked to the CDM community. Below are some collaboration barriers that are being discussed in the CDM community. • At an airport where demand exceeds capacity, how should the capacity be distributed among the demand? In the CDM arena, this was solved by the technique called RBS. In short, for scheduled carriers, RBS demands that the rationing of capacity is based on the original sched- ule of operations, not the current projections of demand. Because the schedule is fixed and published well in advance, the consequences of operational decisions do not affect the allo- cation, allowing operators to focus on making the right decisions for efficiency rather than attempting to game the system. Research is ongoing to determine suitable priority for non- scheduled operators. • How do we compute flight readiness? Real-time flight readiness is needed to properly control taxi queue lengths for departing flights. If various flight operators calculate flight readiness in a different manner, then that variable will impact proper queue of taxiing aircraft. Gener- ally every departure notification of readiness to taxi to ATC (or some local control entity) is only known real-time. If an airport is going to control and allocate taxi queues, then flight readiness must be predicted, so the operator is required to indicate this readiness prediction. For each operator, this readiness indicator might be different. To simplify this readiness state, ACDM efforts plan to utilize a readiness time, the predicted EOBT, and not a readi- ness indicator. • How should arrival and departure runway configurations be utilized by ATC? ATC can apply a preference based on demand. For example, for dual-use runways—those utilized for both arrivals and departures—increased spacing between successive arrivals provides more capa- bility for departures to utilize the same runway. In European ACDM efforts, this utilization is left to ATC with the understanding that the utilization will capture the maximum amount of capacity instead of favoring one operator. NASA is conducting considerable work in this area of airport configuration management and the idea that configuration could be predicated

Challenges and Limitations to Implementation 35 and prepared for by a real-time prediction of all actual demand, including adjustments for the impact of ATC metering and other FAA traffic management initiatives. Examples of some barriers that are particularly relevant to airports are included in Table 7. Metrics Usage In previous sections, we have discussed the value of metrics and the accountability and increased collaboration that occurs from their proper use. It must be stressed that metrics are not to be used to judge others. As previously illustrated, the ACDM projects or tests in the U.S. have shown that metrics of a particular project, determined by the ACDM effort, are a most valuable asset to support and improve collaboration. These same projects or tests have shown that metrics are not used to “grade” an operation, but to support and recognize the aim of future collaboration. As one ACDM participant is quoted, “We don’t keep score.” If this aspect is not continually stressed, especially in the early stages of ACDM, metrics will be used improperly and become a barrier (e.g., comparing one operator’s on-time percentage to another’s, rather than the on-time percentage for the entire airport operation). Appropriate metrics will vary widely, depending on the nature of the ACDM project being undertaken. For example, a departure metering effort will likely focus on departure queue lengths, gate hold times, and information quality. In contrast, a de-icing program may focus on different metrics like time spent waiting for service, time spent on the de-ice pad, and a count of flights experiencing cascading delays. The following depicts a list of possible metrics. The list is not intended to be complete, but to stimulate ideas for measuring the benefits of ACDM projects. • Taxi-out time—length of time from ready for gate departure to actual runway off time. This metric could be subdivided into ramp delays and airport movement delays to source the root cause. • Number of departure flights queued (waiting) at departure runway and time in queue—this could measure the fuel burn and associated environmental impact of departure queues. • Actual number of flights departed during a certain period—measures departure efficiency. • Length of gate hold for departure queuing—if departure efficiency is at a high level, length of gate hold with engines off measures associated fuel savings and reduced environmental impact. Airport Relevant Barriers Possible Mitigation ACDM Implementation Cost: Includes equipment, software, additional staff costs, and training Even though not required, this is the advantage of performing Benefit-Cost Analysis (BCA) detailed in Chapter 6. These include reduction in emissions. airport brand enhancement, improved customer experience, reduced taxi-out delays, handling of flight diversions, etc. Introduction of the ACDM Philosophy to Airport Staff: Includes new procedures and tasks, and acceptance of new working relationships by airport staff CDM history has shown that these barriers will arise. Public commitments by airport senior staff, collaborative development of program goals including means to measure success, and an active feedback loop to address concerns are all extremely important in mitigating these type barriers and should receive robust attention. Modification of Contractual Agreements: Includes those with vendors/contractors performing ramp control functions Past experiences in times of airport construction that required process changes to implement mitigation procedures, these type barriers usually did not exist if they were part of the ACDM process that developed the changes. This reinforces the need for all stakeholder inclusion in the process. Table 7. Other barriers to ACDM implementation.

36 Guidebook for Advancing Collaborative Decision Making (CDM) at Airports • De-ice wait times, number of flights de-iced multiple times, times to actually de-ice, and actual number of flights de-iced—measures de-ice efficiency, queuing efficiency for de-ice, and proper demand for de-ice resources present. • Time arrival flights waiting for a gate—measures gate usage efficiency. • Length and wait times for CBP and TSA functions. • Aircraft in time at the gate to baggage off load completed. • Flight arrival and departure times during period where passenger comfort services are not available. • Number of missed flight connections. • Number of missed baggage connections. • Metrics derived from the above: – Aircraft operating costs – Fuel consumption – Passenger travel time – Greenhouse gas (GHG) and noise emissions. Proprietary Information The proprietary barriers fall into two categories: security and economic. For example, in the security arena, if a display of highly accurate prediction of aircraft position were available, it could be utilized for targeting of a specific flight by terrorists. In the economic area, could it be utilized to gain a competitive edge or illustrate a competitor’s weaknesses? These concerns are not new, but in any CDM effort, including ACDM efforts, all concerned must understand these barriers so that they do not destroy collaboration. ACDM participants must be aware that non-participants within their own company might want to use ACDM data for non-ACDM purposes. Regulatory Issues Concern for compliance with federal regulations (e.g., FAR Part 117, FAR Part 139, TSR 1542) is always very important to flight operators. For example, FAR Part 117 is the regulation that governs pilot duty time. Since one ACDM concept is holding aircraft at the parking gate to con- form to a scheduled taxi time to control runway queue length, concerns might be expressed that this departure metering from the gate might influence pilot duty time issues. However, initial implementations have suggested that ACDM can actually help ensure compliance, as ACDM efforts and metrics make these predictions and conformance much more manageable to opera- tors, especially at a location where departure readiness is forecast. This allows better management of the need to augment or change crews. The ability to forecast and act on these issues, rather than becoming subjected to them without foreknowledge, could be very valuable to flight operators. Similarly, other relevant regulations may actually become easier to approach in the presence of a developed ACDM program (e.g., impact of the three-hour tarmac rule). Scalability and Applicability of CDM As discussed, ACDM efforts will be different at different airports. The local ACDM commu- nity will address the issues that are present at that airport, so applicability is automatic. European ACDM efforts have proven ACDM scalability from applications at major hubs (Paris and Frank- furt Airports) to application at mid-sized airports (Brussels and Dusseldorf). Can ACDM efforts be scaled to large hubs, medium or small hubs, or military/general aviation airports in the United States? The answer is not only yes, but also in certain locales it is going to

Challenges and Limitations to Implementation 37 be needed at smaller airports to allow ACDM efforts to operate at the large or main airport. For example, in the New York area, all departures from all the airports (Newark Liberty Inter national, LaGuardia, John F. Kennedy International, Teterboro, Westchester County, etc.) utilize the same departure fixes. To properly allocate the departure resource, departure demand at all of these airports is needed by ATC, not just the major airline passenger airports. The further in advance this demand is known, the better the allocations can be distributed. Another question often asked is whether smaller airports with limited resources can imple- ment ACDM without sophisticated automation or decision support tools. In other words, is data alone sufficient? The answer to the question is a theoretical “yes.” Theoretical in that implemen- tation of ACDM concepts on this scale has never been attempted and actionable data probably will not be available real-time, but action on analysis data could be useful. Table 8 depicts some possible examples of “data” applications and the attributes of that data. Please note that none of the FAA/Industry CDM data products are utilized, as that requires some type of IT attribute to obtain this data. A prime source for such aircraft movement data might be the Aviation System Performance Metrics (ASPM) database. ASPM is an FAA database of the National Airspace System, a part of FAA Operations & Performance Data and although password protected would generally be available to airports via web access after coordination with the FAA facility. Departure Reservoir Coordinator The CDM Surface Concepts Team concept of an airport Departure Reservoir Coordinator (DRC) has only been put into practice at JFK International Airport. The full ACDM applica- tion of this DRC concept will generate issues that will need to be addressed due to differences in airport attributes. One issue that has not been addressed in the New York area is that departures from all New York airports utilize the same departure “gates” for transitioning to the en route portion of flight. Since the same departure “gates” are utilized, how will the DRCs at JFK, EWR, and LGA coordinate with each other when staging departure flights to the same departure “gate”? The question has been discussed and is expected to be addressed in the future as the FAA TFDM program matures. Problem Analysis Data Attribute Possible Source Taxi-out times Time elapsed from gate departure (OUT) to actual departure (OFF). FAA ATCT data often only includes Movement Area taxi time. 1. Flight operators 2. FAA ASPM data Taxi-out times for a certain subset of flights, e.g., International or Inbound to a major hub. This to identify issues for a specific set of flights so that actionable times will not be lost in the larger average. Time elapsed from gate departure (OUT) to actual departure (OFF). FAA ATCT data often only includes Movement Area taxi time. 1. Flight operators 2. FAA ASPM data Gate availability issues Do arrival flights have to hold in either the Movement Area or ramp area waiting for an available gate? 1. Flight operators 2. Airport operations TSA wait times Time entering TSA queue to clearing from it 1. TSA 2. Airport operations Passenger services Number of departure flights and passengers after service/vendor hours 1. Flight operators 2. Airport operations Table 8. ACDM applications without IT tools.