Autonomous vehicles and toll roads: Existential threat or critical opportunity?
Two themes are common in conversations about connected and autonomous vehicles (CAVs): that widespread autonomy is “just around the corner,” and that “it will change everything.” The former tends to be followed by realizations that this “corner” is actually further away than we thought. The latter tends to focus on expected upheaval in the automotive industry due to changes in vehicle design, manufacturing, and ownership. But what about the infrastructure that transports these vehicles? Particularly for roads with a P&L — tolled roads, congestion zones, managed lanes, etc. — what effect will CAVs have on their revenues, and what opportunities do these changes create? Whereas other disruptions (the Great Recession, COVID-19) produced a dramatic initial shock that attenuated over time, CAVs will gradually and inexorably become more prevalent over the years to come.
Ratings agency Fitch recently published a report on precisely this topic: The Effect of Automated Vehicles on Toll Roads. Fitch assessed the effect of CAVs on total vehicle-miles traveled (VMT) and explored how these effects would differentially impact various types of road infrastructure. Fitch finds that, unsurprisingly, total VMT will likely increase because longer journeys are less onerous to vehicle users who can use the time for productive activities like reading, sleeping, or working. Empty vehicles, impossible in a pre-autonomous road environment, create entirely new journeys. This increases demand for road assets generally, and may also increase congestion if the increase in usage outweighs gains to capacity due to improved CAV navigation. For monopolistic road assets, including many toll bridges, tunnels, and inter-urban highways, this would mean higher revenues.
Unfortunately, the good news for toll roads ends there. For the same reason that CAVs increase VMT, they fundamentally change the calculus of route selection by reducing the “value-of-time” saved by a faster route: if vehicle users can use their commuting time productively, why pay a toll to ensure they get to their destination faster? Monopolistic systems may be insulated from this due to the lack of alternatives, but managed lanes that compete directly with free (congested) routes will suffer. Widespread, advanced autonomy is far away — Fitch puts it at a decade or more. But this remains a serious problem for road assets that were financed with 30–50 year time horizons, using assumptions that no longer hold true. Credit ratings for road assets will fall, raising the cost of financing and making many projects unbankable, if nothing is done to address this.
Crucially, however, Fitch missed one key point: commuting time can’t be used productively if a vehicle user must constantly monitor the road, ready to intervene at a moment’s notice. Therefore, their argument about “value-of-time” really only kicks in at the highest levels of autonomy: L4 within CAVs’ operational design domain (ODD), and L5.
Herein lies the opportunity for managed lanes, in particular: to extend CAVs’ ODD by providing an environment ideally suited to L4 autonomy. The routes that compete with managed lanes are inherently hostile to L4 autonomy: congested, chaotic, and full of “edge cases” that will confuse the computer and require a human to take control. By contrast, a managed lane is free-flowing, predictable, and well-controlled. Managed lanes must invest in technological infrastructure to support CAVs, empowering them to maximize the benefits of autonomous navigation and create value for their users. This turns Fitch’s argument about value-of-time on its head: by creating an environment where time is valuable, managed lanes can convince users to pay their tolls.
To extend their ODD in this way, CAVs require specific capabilities from the road. They need a constant stream of real-time, reliable information about all the vehicles in the road — not just connected vehicles, but also legacy ones that lack V2X capabilities. They need to understand risks, events, and traffic patterns beyond their sensors’ line-of-sight. They need feedback on their own micro-navigation (e.g., in-lane position) to calibrate their sensors and to extend ODD to adverse weather and lighting conditions, such as snow. And they need these data to be consistent, unbiased, and user-agnostic.
To answer this need, a number of companies — both new startups and forward-thinking incumbents — are creating solutions that allow road infrastructure to actively support the edge cases of autonomous mobility. The industry has not yet settled on specific standards and practices, and so it is essential that managed lanes act quickly and decisively to influence its development:
- Invest upfront to trial new, unproven solutions to determine what best suits the needs of CAVs — and to design a viable commercial model to support them.
- Next, deploy these technologies “ahead of the curve,” creating early incentives for CAVs to use them and thereby influencing automotive OEMs’ R&D decisions.
- Finally, continuously seek technological and commercial innovations that will position managed lane operators as key value-enablers, not simply passive infrastructure.
Managed lanes face an existential threat from CAVs, but they also have a unique and fleeting opportunity. By investing in technology that supports the edge cases of autonomy, roads can continue to play a central, value-creating role in tomorrow’s mobility ecosystem.
Valerann’s Smart Road System is designed specifically to support the edge cases of L4 autonomy. Using IoT sensors and machine learning, we measure and map the trajectory of every vehicle on the road in real-time, regardless of whether the vehicle is connected or non-connected. We identify events immediately after they happen, and even predict them in advance using the massive (and growing) quantities of data that our algorithms ingest daily. In doing so, we expand CAVs’ ODD in at least five tangible ways:
- Providing lane-by-lane traffic flow information that allows CAVs to optimize their speed and position for safe lane-merging
- Identifying risks and events beyond the line-of-sight of a CAV’s sensors, enabling safer and more predictable responses
- Offering a continuous stream of data that is not affected by weather or light conditions
- Providing vehicles with real-time feedback about their in-lane location, velocity, and acceleration
- Reducing the computational burden placed on fast-moving vehicles by conducting our own analytics at the edge and in the cloud
Valerann has deployed its system on the largest motorways in the US, UK, Spain, and Israel, and has partnered with Jaguar Land-Rover and Bosch to develop a data marketplace for autonomous vehicles. Learn more at valerann.com!