Enhancing Active Transportation and Demand Management (ATDM) with Advanced and Emerging Technologies and Data Sources

Chapter 6. Operations and Maintenance Considerations

This chapter provides insight into the considerations for operation and maintenance (O&M) as it relates to emerging technologies and data sources. Routine O&M maintenance will be different due to the very nature of these technologies as will other areas such as cybersecurity and data privacy, performance monitoring and maintenance, enforcement, O&M costs, and future proofing.

6.1 Routine Operations and Maintenance Issues

Once the new and/or enhanced ATDM solutions go online, there are a host of O&M issues to deal with. Some items for consideration include performance monitoring, routine maintenance, enforcement, and costs. Moreover, many of the emerging technologies and data sources discussed in this Reference require critical O&M considerations.

Some investigations of active transportation and demand management (ATDM) O&M issues have already been published. One example is the ATM Implementation and Operations Guide, which provides a full chapter on O&M.^{(17, 37)} Some O&M principles discussed here (e.g., necessary data, levels of automation, common performance measures, typical O&M costs, post-project evaluation) are likely relevant to the newer technologies and to the broader ATDM scope. The ATDM Lessons Learned report described numerous real-world implementations, some of which cited maintenance as a significant dis-benefit.⁽⁶²⁾ The technical brief on Data Needs for ATDM ⁽¹¹⁾ discussed the data needs for monitoring and evaluation. The brief says that, although ATDM operations objectives are inextricably tied to performance measures (e.g., travel time, delay, planning time index), supplementary data should also be collected to understand the impacts of exogenous factors (e.g., price of fuel, unemployment rates, other highway improvements, work zones). The Guide for Highway Capacity and Operations Analysis of ATDM Strategies: Analysis of Operational Strategies under Varying Demand and Capacity Conditions ⁽²³⁾ highlights the importance of nonrecurring event data and of forecasting the various percentile outcomes (e.g., expected 80th and 95th percentile worst performance).

Cybersecurity and Data Privacy

Some of the emerging technologies and data sources present some unique cybersecurity and privacy concerns. Many of today's ATDM systems are relatively "closed" from the standpoint that they communicate with sensors, devices, and external systems that are on the agencies' private network. Some of these newer technologies and data sources are connected with outside "uncontrolled systems" and in their nature may have some data privacy considerations. The particular ones that should be noted include:

• Crowdsourced data feeds.
• Cloud computing.
• Big data.
• Connected and autonomous vehicle data.
• Video analytics.
• Bluetooth/WiFi sensors.
• Internet of things (IoT).
• Global positioning system (GPS) and automated vehicle location (AVL) data.
• Mobile sensors.

For enhancing ATDM through emerging technologies and data sources, many of the primary O&M considerations involve management and security of data. The FHWA Reliability Data Guide's section on data ownership and maintenance presented a sample list of fundamental considerations likely to govern O&M levels of effort and expense: ⁽³⁰⁾

• Who will pay to collect, store, and share the data?
• Who (if anyone) can sell the data, and to whom may it be sold?
• Are there any privacy issues in the data that must be addressed (e.g., MAC addresses stored from collection of Bluetooth sensor data)?
• Who is allowed to access the data, and what data may they access (all of it? only a subset?)?
• What purposes are the data allowed to be used for (e.g., if they are collected for analysis purposes only, could they also be used for enforcement?)?
• What data use agreements are in place and must be adhered to use the data (e.g., not allowed to store or disseminate information)?

CAVs are a key quadrant of emerging technology that gain a lot of attention as it relates to cybersecurity (access to a vehicles bus) and data privacy (data that can be potentially used to track vehicle locations). Some ATDM solutions (e.g., multi-modal intelligent signal systems, integrated dynamic transit operation) are specifically designed to exploit CAV technology. Archiving, documentation, and analysis of the resulting data may be critical to the success of such deployments. A presentation entitled "Sharing Connected Vehicle Data on the Research Data Exchange (RDE)" addressed some of the challenges and opportunities associated with high-volume multi-source data from CAV, connected travelers with mobile devices, and other sources.⁽⁴³⁾ Figure 23 illustrates some of the O&M activities needed to enable CAV benefits.

Data security is a key component of data archiving. Vandervalk et al. suggest that working copies of databases maintained on primary servers be replicated in compressed formats at remote sites.⁽⁵¹⁾ These daily backups, which do not reflect a binding legal requirement, ensure that the archive service can be rapidly returned to operation with no significant loss of data if a copy of the database is lost. Both the primary database server and the backup storage are located in climate-controlled machine rooms with uninterruptible power supplies and generator backup power, preventing data loss or gaps in data availability due to power outages. The working copy of the database is stored on a redundant array of independent disk devices, providing error detection, redundancy, and the ability to rebuild missing data upon device failures. Finally, hardware maintenance and security updates are provided for all computer systems by experienced systems administration personnel.

The Real-Time Data Capture and Management State of the Practice Assessment and Innovations Scan addressed issues related to data capture, data management, archiving, and sharing collected data to encourage collaboration, research, and operational development and improvement.⁽⁴⁴⁾ The scan covers five industries: aviation, freight logistics, internet search engines, rail transit systems, and transportation management systems. The scan documented best practices in several areas which are outlined here for reference. Please note that the best practices are strictly voluntary and they are not legally binding. The scan documented the following best practices for access, security, and privacy:

• Generally, the holder of the data controls access to them. Within the transportation and logistics community, this access is carefully controlled.
• There are systems in place that ensure that data can be accessed only by the intended people and only to the degree that they need it. The type of data used by the transportation and logistics industry makes it extremely sensitive, with disastrous consequences for business if accessed by persons with malicious intentions.
• Usually, data access is password-protected, and the following is true:
  • Because data generated within the logistics systems are often financial, strong encryption is placed on such data when they are sent.
  • However, several applications can retrieve aircraft and vessel tracking data, often with other identifying information. The security clearance or password protection to access data through these applications is often minimal.
• The protection of data sources is extremely important. In the search engine industry, it is so heavily protected that there is not even disclosure of how exactly it is protected.

The scan documented the following best practices for data storage and backup:

• Frequent backups and off-site storage are typical.
• Preventative maintenance should be performed regularly.
• Careful consideration should be devoted to determining how much and for how long data should be stored. In aviation, for instance, data are kept for a relatively short timeframe because the need is for real-time rather than historical information. At the same time, data can be available for revision if there is an incident to investigate.

The scan documented the following best practices for O&M:

• Deployment should be started on a reasonable scale, such as implementing in a small geographical area or using easily manageable data.
• Multiple servers should be used to distribute real-time loads. Several technologies enable this load distribution.
• It is important to consider determining the needed resolution or granularity of the data. This may vary depending on the context and use of the data. Specific examples include the following:
  • In the logistics and retail industries, inventory data are refreshed every minute in several stores. This is used to support restocking and also to monitor trends.
  • In the search engine industry, data generally go through a 24-hour refresh cycle, staying fixed between cycles.
  • In the aviation field, data are mostly retrieved as fast as possible to enable incident prevention.
• It is necessary to determine what is critical to communicate and what is not. For instance, railroad and airline alert systems only collect the necessary data that can alert an operator of a particular problem.

The scan documented the following best practice for critical failures:

• A common issue is that correcting a problem is often dependent on a single person, meaning its solution depends on the person's availability. It is, therefore, important to have staff available around the clock to solve potentially catastrophic failures. The higher labor cost is a necessary expense if the system needs to be highly available at all times.

Performance Monitoring and Maintenance

As discussed in the ATM Implementation and Operations Guide, performance monitoring is an integral part of the active management cycle, which was illustrated in chapter 1.^{(17, 37)} Performance monitoring is an ongoing internal process where system conditions and performance are examined and evaluated through data collected through interfaces with devices/equipment installed in the field as well as data feeds such as probe or crowdsourced data. Performance monitoring provides the data needed in the decision-making process. Deployment agencies use data and performance monitoring during strategy activation. Activations typically fall into one of two types of processes, automated systems and manual systems, which vary greatly in the strategy activation thresholds used. In some cases, a hybrid automated-manual process is used.

ATDM deployments are often part of an agency's overall transportation management system, which is a complex, integrated amalgamation of hardware, technologies, and processes. System maintenance includes replacing worn components, installing updated hardware and software, tuning the systems, and anticipating and correcting potential problems and deficiencies. Maintenance includes the development and implementation of action plans for responding quickly, efficiently, and orderly to systemic failures, as well as having infrastructure and procedures for measuring and monitoring maintenance activities. An agency's maintenance strategy can dictate system design and must be considered in the planning phase to ensure that it will have the personnel or financial resources to adequately maintain the ATDM solution.

The ATM Implementation Guide provides more specificity on performance monitoring for specific solutions.^{(17, 37)} For example, the guide lists a of sample agencies who implemented automated thresholds for ramp metering activation. Such thresholds were typically a combination of mainline occupancy, mainline volume, ramp queue length, and/or ramp storage length. Threshold values may be updated periodically (i.e., monthly, quarterly, or annually) based on continued assessments of system performance. Because the effectiveness of short-term and long-term performance monitoring depends on the quality of data collected, having access to quality data is an important element of performance monitoring. Emerging technologies and data sources can augment performance monitoring in active traffic management (ATM), active demand management (ADM), and active parking management (APM).

Active Traffic Management

Many strategies activate at specific traffic congestion levels to prevent, mitigate, or delay the onset of traffic bottlenecks. To accomplish this, performance monitoring is critical to determining when strategies should activate. Some predictive strategies employ combinations of historical and realtime data. Bayesian models use real-time data to improve the accuracy of predictions based on historical data. Machine learning (ML) methods are able to fine-tune and improve these prediction algorithms over time. Beyond monitoring traditional metrics such as flows, speeds, and densities, inclusion of non-recurring event data (e.g., weather, incidents) can help to explain performance and improve predictions. In summary, ATM performance monitoring can be augmented through (1) fusion of historical and real-time data (e.g., Bayesian methods), (2) explicit use of non-recurring event data (e.g., weather, incidents), and (3) ML.

To augment the ATM solutions as described above, emerging technologies can help to improve both the quality and quantity of available data, especially when compared and/or combined with more traditional forms of data. For example, probe data are now used extensively to obtain traffic performance metrics. Social media provides a source of incident data.⁽⁵³⁾ Dynamic message signs (DMSs) can display optimized sets of driver information based on enhanced data sets and learning methods. Some urban data sets, including both traditional traffic sensors (e.g., loops, cameras) and cutting-edge sensors (e.g., Bluetooth, GPS probe, parking), have been archived for a decade. These rich data sets allow learning of traveler behavior and in-depth understanding of non-recurrent traffic. They can be applied directly to predict traffic impact of planned and unplanned incidents and provide real-time decision making for traffic operations.⁽⁵⁵⁾ Figure 24 illustrates an example of this ATM augmented by emerging technologies.

Regarding O&M issues for ATM, quality control has been a concern for probe data. Agencies may wish to monitor the proportion of missing data over time within probe data sets. Along surface arterials, accuracy of probe data may be compromised by the proximity and influence of traffic signals. Also, the proportion of vehicles surveyed within probe datasets changes over time and may become unacceptably low in some areas. Beyond probe data, weather and incident data may be less accurate on certain roadway segments. In summary, cutting-edge sensors and learning algorithms may require annual audits, calibration, and repairs, just as traditional sensors would.

Active Demand Management

Under ADM, traveler choices can be influenced by access to the right information. In many cases, this information can be improved and/or optimized by emerging technologies and data. In the Stanford CAPRI study described later in this Reference,⁽¹³⁾ commuters were successfully incentivized to re-distribute their departure times in ways that were far more efficient for the overall surface network. In addition to an innovative incentives structure, the program used emerging technologies such as radio frequency identification (RFID) sensing, smartphone applications, and social media. In theory, the traffic flow incentives and monetary awards could be further optimized by a fusion of ML and performance monitoring.

In another example, automatic passenger counters (APC) and AVL data have been used to optimize the supply and demand for public transit. The essential idea is to fully utilize the big data in public transit to provide travelers fine-grained, customizable information regarding transit service performance (efficiency, reliability, and quality). This information can be distributed via DMSs at transit stops or on smartphone applications. An example of such information is shown in figure 25. Moreover, transit providers can monitor day-to-day transit service, and can monitor how transit users respond to information. They can develop a better understanding of travelers' preferences on efficiency, reliability, and quality of transit service, as well as their modal choices. Big data and data-driven behavioral models facilitate agencies' decision making (such as scheduling). Effective information provision, along with data-driven scheduling, holds great potential to improve the service performance and travelers' riding experience.

Regarding O&M issues, RFID, APC, and DMS devices may require physical maintenance and repairs. Similarly, social media data and smartphone applications software may require periodic technical support and software updates. For example, if a social media vendor releases a software update that modifies its input-output format, any data mining tool used to access the social media data will require corresponding updates and software patches.

Active Parking Management

Parking can consume a significant amount of the trip costs (time and money) in urban travel. As such, it can considerably influence travelers' choices of modes, locations, and time of travel. Advanced performance monitoring via smart sensors, wireless communication, social media, parking meter transaction data, and big data analytics offers a unique opportunity to tap parking's influence on travel to make the transportation system more efficient, cleaner, and more resilient.^{(56, 57, and 58)} Essentially, these emerging technologies and data sources may facilitate:

1. Changing day-to-day behavior of all commuters through day-to-day travel experience and/or online information systems.
2. Changing travel behavior of a fraction of adaptive travelers on the fly who are aware of time-of-day parking information and comply to the recommendations.
3. Influencing market prices of privately owned parking areas through a competitive parking market.

Figure 26 illustrates the APM-centric active management cycle.

Regarding O&M issues for APM, the quantity and location of available parking spots is constantly changing. Any cyber-physical system based on parking data will require constant updates. Social media platforms and parking meter transaction data are subject to software updates. Wireless communication technologies are not static. To combat these instabilities, there is a need for human audits, post-deployment monitoring, and dedicated O&M budgets. ML, which represents an automated form of audits, is also highly advisable to complement and support any human audits. Depending on the success of the automated monitoring system, the expenses associated with having humans in the loop may dissipate over time as the human effort decreases accordingly.

Enforcement

Law enforcement opportunities and capabilities are evolving along with the new technologies. This has ramifications for ATM and APM. The typical ADM applications (e.g., dynamic ridesharing, on-demand transit, dynamic pricing, predictive traveler information) may not have significant enforcement concerns.

One prominent issue is traffic violations involving CAVs. These vehicles could potentially be programmed to avoid speed limit violations, parking violations, red light running, and usage of closed lanes. An increasing market penetration rate of such vehicles could significantly reduce the burden of law enforcement. However, there are many privacy issues associated with CAVs. The driving public will not necessarily accept unfettered control of their driving behaviors and identification of their whereabouts. Therefore, the enforcement impact of CAV technology remains uncertain.

Active Traffic Management

Adaptive ramp metering and traffic signal control strategies can now be deployed on freeways and arterials. Red light running and speed limit violations are two enforcement concerns related to traffic signals, regardless of deployment location. While CAVs may help to prevent such violations, automated traffic signal performance measures may reveal such violations from conventional vehicles and drivers. Automated vehicle or license plate identification may streamline the enforcement process.

Two other ATM solutions affect which lanes drivers may use. Dynamic lane and shoulder use allow for temporary opening or closing of travel lanes in response to increasing congestion or incidents. Dynamic junction control prioritizes the critical roadway to minimize the impact of merging and diverging maneuvers. In these cases, illegal use of a closed lane becomes an enforcement issue. While CAVs may help to prevent such violations, automated vehicle or license plate identification may again streamline the enforcement process with conventional drivers. The automated identification may be more technologically challenging along lengthy (i.e., up to many miles long) dynamic lanes or shoulders, on which the violating vehicle location is much less certain than at traffic signal stop-lines or stop-bars.

Finally, dynamic speed limits may change based on road, traffic, and weather conditions. Speed limit violations are possible. A combination of technologies is needed to identify vehicle speeds (not necessarily in real-time) and identify the violating vehicles or drivers. Again, the frequency of speed limit violations may potentially be reduced by CAV technology.

Active Parking Management

APM solutions can dynamically modify parking spot permissions, prices, and availability. Smartphone applications (e.g., figure 27) that facilitate parking spot payments from remote locations should interface with any APM solutions that may be in effect to prevent unnecessary violations. Automated vehicle or license plate identification may help to reduce law enforcement labor and costs. The frequency of parking violations may potentially be reduced by CAV technology, which can be designed to avoid such violations. However, when APM solutions are in effect, infrastructure-to-vehicle communication may be needed for CAVs to avoid such violations.

Costs Associated with Active Traffic Management

Costs associated with ATM merit consideration long before implementation. Cost data are available through various tools and resources (i.e., TSMO Benefit Cost Compendium,⁽⁶⁰⁾ Intelligent Transportation System [ITS] Joint Program Office [JPO] online database,^{(47, 61, and 62)} Tool for Operations Benefit Cost Analysis, Operations Benefit/Cost Analysis Desk Reference to support Tool for Operations Benefit Cost Analysis).⁽²⁷⁾ Some of the major costs include capital investment into infrastructure, technology, long-term operations, maintenance, and upgrades of the system over time. Usually, the Federal Government provides funding for ATM solutions and projects for both initial construction and ongoing costs. Various Federal Government initiatives provide opportunities for funding ATM solutions individually or along with other activities or projects. State and local level funding are also available. Some ITS infrastructure installations and integrations that may qualify for Federal funds are listed below:

• Emergency services.
• Incident management systems.
• Electronic toll collection systems.
• Electronic fare payment systems.
• Freeway management systems.
• Transit management systems.
• Traffic signal control.
• Regional ITS architecture development.

Emerging ATM technologies also require hardware, software, and software integration. Therefore, the long-term support of software, including routine checks and modifications, is a consideration. The integration of CAVs into ATDM solutions will involve implementing advanced infrastructure, which needs to be properly maintained for a longer lifespan.

6.2 Future Proofing

A big challenge with technology systems is how to keep them current so they will not need to be constantly updated (often at high costs) as technologies advance. For operational systems such as ATDM, these systems do not need to be quickly obsolete and/or require a replacement shortly after the system is installed. It is difficult in today's quickly moving world of evolving technology to have systems such as these to be truly future-proof, but there are steps that can be taken to help ease the impact of evolving technology. These steps, which are not required, include the following:

• Perform technology maturity and evolution assessment before implementing a new technology or data source. For example, if a new technology is being investigated, key questions to ask are: (1) how many companies or vendors are doing it, (2) how prevalent do the technology industry experts see the future vision of the technology, and (3) what are the size of the companies and systems that use it, etc. If there is an interesting new technology, but only one of two firms sell the product, this may be a flag for its long-term longevity of maturity.
• Select technology options that offer flexibility, scalability, and consistency. For example, cloud computing is viewed by many as very future-proof because upgrades and updates can be invisible to the user and the system scales seamlessly. These types of environments will keep up with current technology.
• After implementation, perform regular change control meetings, typically with a change control board. This will allow agencies to predict future changes and plan for them in advance. For example, a software product vendor may announce that the current version of the product is being discontinued in a year and an upgrade will be required. This will allow the impacted agency to assess the effect including costs and plan in advance. It will also allow agencies to potentially look at alternate replacements if the upgrade path is too disruptive or costly.
• Build a future-proof information technology (IT) plan. A future-proof IT plan includes taking into consideration future IT trends, changes in material acquisitions, changing price models, new and emerging options for system upgrades, and which upgrades may be the most efficient.
• Create continuing education programs that allow staff to constantly learn about new technologies and trends so they can be brought forward as part of future-proofing exercises.
• Work to have a good working relationship between transportation departments, transportation operations and IT departments. Through working together on items like continuous technology improvement plans, the pitfalls of technology obsolescence can be avoided.