What is Integrated Mobility? There are different definitions to this suddenly hot buzz-phrase, but the general concept is to provide comprehensive connectivity between different travel modes to provide users with the quickest, safest, most reliable trip, as Glenn Havinoviski explains
The concept of Integrated Mobility has emerged as an integral element in the future Smart City, where data on all aspects of the transportation system, sustainable technologies and automation will play a major role. But many of these concepts have been bandied about for a decade, particularly on an international stage, whether by governments or large-scale integrators and technology companies. In the US, regional transportation planners have used terms such as “multi-modal” and “seamless” transportation services, and the USDOT’s Integrated Corridor Management philosophy has been oriented toward modal and route shifts based on real-time conditions, but focused on defined urban corridors. Trends have occurred in the last decade that have introduced new wrinkles into the Integrated Mobility concept. In particular, they alter what many had understood to be traditional public sector and private sector roles. In this article, I offer some personalized observations and perspectives, plus some thoughts on how the ITS world has evolved and how concepts such as autonomous vehicles and Mobility as a Service (MaaS) can become an integral element to Integrated Mobility.
A Changing Transportation Landscape
Having spent two years working overseas, I came home in 2014 to a US transportation landscape that had changed. Yes, the battles for transportation funding at the Federal level have continued as they have for the last decade, although relieved somewhat by the Federal FAST authorization by Congress in December, the first all-new transportation spending bill in ages. And as system operators and consultants, we have enjoyed a menu of projects involving traffic signal systems, 511 systems, ITS architecture, Active Traffic Management, Integrated Corridor Management, regional traffic management centers, transit signal priority and managed lanes. There are many metropolitan areas that could still benefit from these solutions, so in that regard, our work is far from done. But these things rarely capture the imagination of the public the way driverless cars and electric vehicles do, and that the arranging of local trips instantly on a smartphone has. And then there is the ubiquity of traffic data, available on multiple applications and used widely by the public, without any thought to whether it is “Google data” or a public agency’s traffic sensor data.
Driverless cars are now all over the news and have become anointed by many as representing the future of mobility and many of us know at least one person who owns an electric car such as a Tesla, Chevy Volt, BMW i3, i8 or Nissan Leaf. In much of the US one may go to a supermarket, shopping center, Metro stop, office or apartment parking garage and see electric car charging stations.
Back in 2009, Uber was a start-up, and a trendy way for revelers with their cool smartphones to get home after a night on the town in places where taxis were hard to find, and driving personal vehicles while under the influence had become universally unacceptable. Within just five years, Uber, Lyft, and car-sharing firms such as Car2Go and Zipcar, had changed the landscape of what personal mobility means, to the point where the concept of not owning a car is attractive to numerous people, many of them the all-important millennial population. It has also fundamentally changed our culture, so much of it built around being independent and self-reliant, manifesting itself from a transportation perspective in our love of all things motorized (automobiles, motorcycles, etc), all fed by fossil fuels.
And traffic data is available practically everywhere thanks to the broad array of mapping and traveler information applications available for smartphones, the majority of them privately developed and operated. Many regions also have real-time transit data, including next bus and next train, available on smartphone apps. Many of these applications use standard data feeds from the transit agencies but were developed not by transportation of government professionals, but by students, entrepreneurs, and others with some degree of programming skill.
So in short, over the last decade, we have seen the general emergence of:
• Autonomous vehicles (many of which are electric powered)
• “Transportation network companies” (as mobile ride dispatching and car-sharing companies are termed now) handling trips within urban areas, including “last mile” trips from train stations, airports, etc. traditionally handled by taxis
• Apps such as Google Maps, Inrix Traffic, and Waze, all far more commonly used to get real-time traffic and transit information (and for some apps, also providing data back).
What do all of the above have in common? It’s fairly simple. They were largely done without traffic engineers and transportation planners thinking them up. They didn’t require Federal or State funding or contracting to build. There were no acts of Congress or legislatures needed, no TIGER grants, or earmarks. No “bridges to nowhere” were constructed. They didn’t map clearly into the US National ITS Architecture. They don’t use DSRC at 5.9 GHz, one of the original cornerstones of the long-established and frequently-renamed Connected Vehicle initiative.
What has changed is something that ironically, the founders of ITS (as we now know it) in the early 1990s wanted to happen. That is, the private sector has taken it upon itself to develop applications and services that utilize advanced technologies and communication strategies to improve transportation. The only thing is it hasn’t necessarily been driven by government policy, or by the government’s significant and valuable initiatives. Rather, the private sector identified a market for such services, along with their potential for monetization and revenue, and is now developing that market. Through partnerships with some long-time vendors in the ITS world, such as mapping and GPS companies, as well as automakers who offer their vehicles as mapping, communications and entertainment platforms, the private sector may suddenly become the partner the public sector has needed to make ITS not just a government function, but an essential part of people’s lives.
In all this, there is a fundamental question. To implement the ITS of the future, does the public sector need the private sector more than vice versa? What can the public sector do to create an attractive playing field for all vendors and developers? Is it in developing standards for communication interfaces and data feeds, or wireless infrastructure? We are finding many of the activities by the private sector involve nothing that didn’t already exist thanks to the growing wireless communications infrastructure and development of mobile application environments by Apple, Google and Microsoft.
In fact, the government should take some credit. It is in the actions taken since the 1980s and 1990s to open up GPS technologies and wireless bandwidth as well as opening civilian access to the Internet, that the seeds of our new transportation revolution were sewn. We may have not seen it, because we were focused on how we could enhance the public sector role in operating the transportation infrastructure, managing recurring and non-recurring congestion, and regulating vehicle safety, along with its role in interstate commerce including freight. But while focusing on these public sector activities, more than a few people were scratching their collective heads wondering when and how this was all going to pay for itself. Perhaps in light of all the fundamental changes on how we look at vehicles and personal mobility, this may or may not even be a relevant question anymore. So how does this play into Integrated Mobility?
Integrated Mobility Trends
In 2013, at the International Road Federation World Meeting and Exhibition in Riyadh, Saudi Arabia, there was a very broad and intriguing technical track on “Integrated Mobility and ITS” which I was honored to chair. The final day of the meeting was devoted to conclusions and outcomes from the various sessions held under each technical track. For Integrated Mobility and ITS, they indeed point to a fundamental shift in how transportation agencies and service providers (public and private) need to address public expectations in reducing travel delay, providing convenience and travel choices. The IRF technical track outcomes indicated several key trends in managing and operating transportation services, as shown in Figure 1. They are still valid in 2016, but are manifesting themselves in different ways. Below these trends related to Integrated Mobility are discussed in terms of what they would accomplish, but also how they are impacted by the more current trends related to autonomous vehicles and MaaS.
Person-Trips vs. Vehicle-Trips
In recent years, the performance of transportation networks as determined by metropolitan planning organizations (MPO) has been increasingly focused on Congestion Management Planning activities, focusing on key outcomes which involve individual mobility, rather than vehicles. This is often done through incentives to increase vehicle occupancy as well as providing choices such as rapid bus or rail transit, managed toll lanes and other strategies. The idea is to maximize person-carrying capacity and allow greater mobility with the same or fewer vehicles on the network.
The challenge today is in (a) providing adequate funding to construct, operate and maintain transit services despite their increased usage, and (b) many of the trends associated with MaaS and autonomous vehicle operations focus on individual trips and users, not maximizing vehicle capacity or efficiency for a network. However, autonomous vehicle research is already pointed in a direction where vehicle platoons for passenger and freight travel would be feasible in the future, provided there is a Connected Vehicle component as well to manage travel speeds and flows based on other conditions on the network (congestion, incidents, road work, weather, etc.).
Multiple Sources for Data vs. Single Sources
Data fusion from both roadside and vehicle data sources has been an ongoing trend over the past two decades. Through the early 2000s, the primary focus by State DOTs or other transportation operators was to instrument the road networks to help monitor and manage traffic flow and incidents. Loop, microwave, and machine-vision sensors provided quantitative data, while pan-tilt-zoom CCTV coverage allowed for qualitative monitoring and confirmation of incidents and conditions on the same routes, often between sensor locations. The introduction of the Real-Time Transportation Management Information Program requirements in Section 1201 of the SAFETEA-LU Federal reauthorization in 2005 introduced a requirements to provide real-time information both inside and outside urban areas, creating new challenges for public agencies in supporting national objectives of measuring transportation network performance and identifying the benefits of specific programs and projects.
However, the availability of travel time data from Bluetooth monitoring along roadways (generally more inexpensive to construct than traditional sensors), and the increased use of smartphone-based probe data by private sector firms such as Inrix, Google and Here has introduced a level of coverage across urban and rural areas in the US and internationally that wasn’t economically feasible with roadway ITS infrastructure. Many States such as Virginia now purchase data from private sector entities containing flow information to help manage and monitor performance, although count data and more precise road use and classification information to date still relies on infrastructure sensors. Likewise, tracking bus and rail location provides an opportunity to provide comparative traveler information.
From a planning perspective, trip data from MaaS services such as Uber, if made available to local agencies, would enable transportation planners to understand changes in travel patterns over time. Autonomous vehicle operations introduce an interesting data element. First and foremost, they must rely on information from their immediate surroundings, using sensors which detect proximity to obstacles, nearby vehicles, and lane-keeping algorithms, as well as visualizing traffic control displays. In general, the autonomous systems being developed utilize map-based traffic flow data to generate a travel path of least resistance to a specific destination, and as a rule are not relying on public sector generated flow data.
To assure truly efficient operations and enhance safety, however, autonomous vehicles would greatly benefit from connectivity to roadside devices and to data, eg to use signal timing data and downstream travel conditions to further inform the vehicle operation, reduce sudden stops and react to activities such as road work or lane closures. Thus, as is being emphasized by USDOT and many experts, future autonomous vehicles should also be connected vehicles. While CV-type information could ostensibly be provided from map-providers over high-speed wireless connections independent of roadway-based DSRC, the level of accuracy in communicating the data to the vehicle must be high, there needs to be connectivity from the public agency/road operator to the information service provider, and conversely there needs to be the ability to gather information from the vehicle to the service provider and to the public agency relative to incidents, emergencies, or real-time conditions.
Current applications such as Waze and Inrix Traffic provide manual reporting of incidents and congestion within the app, and subscribers to these services in the public sector can receive this information. Such functionality would be one area that would need to be automated in an autonomous and connected vehicle environment.
Multiple Interactive Systems vs. Single Systems Working in Parallel
To a great extent, traditional “stovepiping” of individual systems (freeway, traffic signal operations, public transit monitoring, and traveler information) had occurred due to institutional responsibilities and resultant liability considerations. Development of the National ITS Architecture and related standards development activities, plus subsequent efforts in Regional ITS Architecture development and ITS strategic planning, have done much to enhance coordination between public agencies and introduce a more regional flavor to Transportation System Management and Operation (TSMO) activities with time. While since the 1990s many adjacent jurisdictions have coordinated signal timing activities, the early efforts were often facilitated by those agencies using common signal system vendor platforms and controllers. Subsequent standardization activities and cooperative systems procurements involving multiple agencies have furthered interjurisdictional coordination efforts. Missouri’s Gateway Green Light program in suburban St Louis and Operation Green Light in the Kansas City area are two such efforts.
Sharing of freeway congestion and arterial traffic data has occurred more frequently as part of regional traffic management strategies, and has been facilitated by the NTCIP-driven standardization of center-to-center communications. Integrated Corridor Management (ICM), developed as a concept over the past decade and implemented in Dallas and San Diego using a combination of real-time data fusion and decision support systems, brings together data from freeways, arterials and transit services for the purposes of managing recurring and non-recurring congestion, and also can serve as the basis for multi-modal journey planning.
Autonomous vehicle operations stand to benefit from the data provided from different systems allowing travel paths to be identified and coordinated with real-time agency information on road work and incidents. Although much of the testing of autonomous vehicles involves communication of probe data from private mapping and traffic information services, the nature of that data may be more reactive than proactive if there is no public data element which may describe events that are or will be occurring on the road network, or other information such as parking availability.
Transportation network companies introduce a wrinkle in terms of travel demand, and that is the use of individual vehicles in lieu of transit to provide the type of door-to-door services that taxis provided traditionally and that private vehicles can do to a limited extent (depending on parking at the destination). Without adequate information on these travel patterns, particularly as such services continue to grow, much of the network management may remain reactive as opposed to proactive. Integrating such MaaS providers as transportation options next to transit, toll express lanes, or conventional road travel is one area of interest, with an eye toward understanding the potential role of transportation network companies in providing first-mile or last-mile services relative to rail or other transit corridor services.
Finally, the provision of pedestrian facilities and bikeshare services has become a major priority in several cities as a sustainable travel option to reduce motorized vehicle traffic in high-density areas. This too, provides a viable option for first-mile and last-mile services.
Pro-Active Management and Information vs. Informing/Reacting
Probably the biggest push in ITS-related TSMO in the past decade has been the emergence of more strategies that provide active management of traffic flow, whether it involves tools such as variable speed limits and part-time hard shoulder running strategies on freeways, adaptive traffic signal control on arterials which varies timings and patterns in reaction to current traffic flow changes, or providing priority for buses and light rail vehicle movements on arterial routes. To a great extent, pro-active management relies on anticipating the likely impacts of current traffic flow trends compared with historic trends, as well as facilitating transit movements when there are significant riders and potential delay on key routes. Congestion-based toll rates are a key component of priced managed lanes and increasingly urban area toll road facilities, increasing the toll in real-time based on downstream flows or on historical travel trends.
Ramp metering was an early example of active traffic flow management, especially when variable metering rate algorithms responding to congestion on the freeway were implemented in areas such as Chicago and Seattle, with further integration of ramp metering and nearby arterial signal operations implemented by Caltrans at several locations in Southern California.
Autonomous vehicles, by definition, will have a fundamental impact on TSMO. On one hand, autonomous vehicles need to react to a plethora of inputs from their surroundings as well as real-time data from a mapping source. At the same time (as discussed above), connected vehicle technologies would enable the autonomous vehicle to have additional intelligence about conditions otherwise unseen by the autonomous vehicle, whether they are informational and advisory in nature (toll rates, parking space availability and location, downstream travel speeds), or whether they provide immediate safety messaging benefits (eg, intersection collision warning).
MaaS offers another wrinkle in pro-active management activities, which is the influence of individual travel on the network. Because such travel is not often predictable, the key influence might be the total percentage of vehicles and personal trips on the network that are being taken, whether through ride services (eg, Uber, Lyft) or through shared vehicle resources (eg, Car2Go, Zipcar). If these types of trips remain a relatively small percentage of all trips (say, less than 5 per cent), it is possible, at least in the short term, that they might not have significant influence on traffic management. However, if the percentage of such services increases, and as is widely predicted such rides and vehicles involve driverless, automated services, then there could be a significant impact, and some degree of cooperation and coordination is needed between the autonomous operations and management of transportation networks. Long-term, there is discussion that fully autonomous networks (100 per cent of all vehicles operating driverless) would no longer need signalization or signage, but that is still, practically speaking, well-off in the future.
In-Vehicle Connectivity vs. Roadside Traffic Control Devices
This trend is focused in particular on the emergence of the connected vehicle and related Vehicle-to-Infrastructure (V2I) communications. This addresses the applications which place key traffic control data and displays within the vehicle, as well as immediate safety messages (intersection collision warning, reduce speed) which in turn can trigger automated features within the vehicle. This outside knowledge assists an automated vehicle to the same extent the same information assists a driver, except that automated corrective actions could be accomplished much faster, reducing or eliminating reaction time. In the medium- to long-term (eg, the next 10 to 20 years), providing in-vehicle advisory and guidance information (traffic data, map data, weather warnings, real-time information messages and route diversions) could reduce the need for dynamic message signs and other devices which may currently provide such information for drivers. However, providing more direct traffic control information (ie, “stop” and “go”) requires a far more secure, direct communications connection with the control devices issuing such control commands.
To a great extent, the seeds of wireless guidance are found in today’s in-vehicle navigation systems that utilize real-time traffic information, as well as the wireless phone applications of Uber, whose driver apps contain dispatch and route information to reach the traveler’s destination and provide the fare information that is debited from the user and credited to the trip, with a percentage going to the driver. Automation of such activities becomes a next step.
Challenges and Opportunities
In providing this perspective on how the concept of Integrated Mobility may evolve thanks to vehicle automation and ubiquitous mobility applications (largely private sector in nature), some thought should be given to the challenges, and conversely, the opportunities we face as the transportation community
Today, we are seeing data – both historic and real time – increasingly used to conduct predictive and adaptive traffic control and freeway management strategies that allows far superior management of existing road assets. At the same time, this data offers the traveler a greater palette of multi-modal choices with much less need for investments in new physical infrastructure. A traveler may drive, share a ride, and make use of an alternative facility for a user fee (for example, managed lanes) or use public transportation. The data may be used to provide real-time downstream information, may manage traffic flow through tools such as variable speed limits and lane control strategies, or can provide the traveler with personalized, alternative mode travel time and cost information through web or mobile “journey planner” applications.
For many public agencies, much of the data being used to support traveler information and manage corridor activities is already being procured from the private sector for entire corridors or for all State highways (including US and Interstate), and in some cases validated using Bluetooth-based roadside monitoring. In general, the procured data is probe data coming from mobile phones in vehicles, although earlier initiatives involved implementation of technologies such as microwave-based detection to collect traffic flow data such as volume and speeds. In some cases, information on incidents and current travel speeds is being corroborated through traffic-oriented social media. Quantitative use of this data can support Real-Time Management Information Program activities as directed through SAFETEA-LU and subsequent Federal authorizations.
Vehicle Information and Privacy Concerns
However, as connected and autonomous vehicles are introduced into the transportation network, the data being captured from vehicles will be far more granular and detailed, involving not just road performance but also more specific data related to vehicle operations and environmental conditions (eg, wiper operations, indicating precipitation, or wheel slippage, indicating road icing).
Although this data may be considered to be government-collected data in the same manner as agency-collected sensor data has for traffic management systems, there is also the potential for direct correlation of the data to an identified vehicle and user. The protection of personal privacy relative to the collection of vehicle and mobile data will likely be of key concern to public policymakers and to individuals. Stripping out data identifiers and any links to financial information associated with a person or vehicle will be critical in this regard.
However, the trust issue does and, in all likelihood, will continue to be a public concern. One option, to allow users to “opt out” of certain types of messaging, may result in less information being made available for operational purposes. Thus, for some routes and corridors, a degree of redundancy involving agency-owned infrastructure-based sensors, commercially-collected probe data, and connected vehicle data, may be appropriate. How this is utilized and how it is archived will be one of the major challenges, as well as opportunities, in the future.
With the increase in data sources from vehicles, commercial data providers, and agency-owned sensors, the amount of data stored will increase exponentially. Although data storage technology and capability has increased tremendously with time, the ability to access, query and utilize the appropriate data in a timely fashion will be of paramount importance. How much is stored and how long it is kept will become of increasing importance, as will data reduction analysis techniques and tools that eliminate data or paths that may not be relevant for particular dynamic traveler routing or management analytics. Latency of data collection (which may be a function of the data collection technologies themselves as well as communications links) may not be as significant an issue as the processing or smoothing of data from different sources.
How Do We Address the Role of TNCs?
Data from TNCs that provide smartphone-enabled ride dispatch, ridesharing, and vehicle-sharing services will be increasingly valuable in assessing trip patterns and managed operations in a dynamic fashion. As many urban residents are noticeably giving up personal automobiles for shared services, transit, or non-motorized travel options, tracking this information will be valuable in developing a plan for developing an Integrated Mobility strategy as well as serving as a basis for assessment and evaluation of the strategy over time. Arrangements between TNCs and local/State agencies for obtaining traffic flow information as well as origin-destination data would be of significant importance to the public sector in identifying investments needed in roads, parking, fixed-guideway and variable route transit services, and electric-vehicle charging locations, among other data.
However, institutional relationships between many TNCs and public sector agencies have been sensitive (in some cases a significant issue as a result of regulatory structures or lack thereof) and many private companies see their information as proprietary in nature.
Further, it is expected TNCs will move toward use of autonomous vehicles and eventually driverless services (whether to provide trip services or to “drop off” a vehicle for the user). Obtaining and storing real-time vehicle monitoring information will be critical in understanding how road capacity, traffic engineering and parking needs will change in the future.
Integrated mobility is still not an absolute reality today, as most travelers rely on either driving or transit or specific mobility applications, as opposed to a mix of the above. (Although park-and-ride services can blend personal vehicle travel for the first mile/last mile with transit or ridesharing for much of the trip to the work or other destination, and then back to the park-and-ride lot where the personal car is used to get home.) Likewise, many public agencies still operate their arterial management, freeway management, and transit operations in relative isolation, even if some data sharing occurs to support real-time traveler information as well as supporting certain areas of coordination such as traffic signal priority. However, many emerging activities that are occurring today, which in turn support the emerging Smart City concepts, do entail the fusion of data, use of information from outside the public sector infrastructure, and use of a mix of public sector and private sector mobility products to achieve a future Integrated Mobility vision.
Glenn N. Havinoviski, PE, is Associate Vice President, Transportation Systems for Iteris, Inc.