ClearRoad FAQ

Starting off with questions from the So You Think You Know Congestion Pricing… webinar

It can be hard to recall events pre-pandemic, but I checked my calendar, and a few days prior to the Bay Area shelter-in-place order, I, in fact, presented in NUMO’s webinar So You Think You Know Congestion Pricing [View the recording there!].

Now that six months have passed since the ill-timed webinar, I’m belatedly responding to many of the great questions from the 100+ in attendance. Going forward, this will serve as a living document for ClearRoad to respond to questions about congestion pricing and road pricing more generally. This was originally published on September 24th, 2020. Questions are lightly edited, grouped, and combined where appropriate:

Congestion/Road Pricing Lessons Learned

What would you say are the biggest lessons learned and key takeaways from cities that have implemented congestion pricing?

• The classic lesson from London & Stockholm is that politicians need to have iron stomachs to withstand pushback and protest up until the moment of implementation. But those making it that far are virtually guaranteed to see a positive public response on the other side.
• Pair implementation with viable transportation alternatives. London increased bus options substantially before turning on the system.
• Don’t include too many exemptions and discounts as they increase incentives to cheat, increase the fee burden for the non-exempt, and generally undermine systems’ success.

What key lessons can cities looking at congestion pricing learn from this new approach?

Below are some relevant lessons learned from ClearRoad’s experience in Road Usage Charging (RUC) and gleaned by states from polling participants:

• Participants ended up, by a substantial margin, preferring reporting solutions that they didn’t have to think about beyond the initial setup — “set it and forget it.”
• Providing multiple data reporting options, e.g., plug-in devices often used in Usage-Based Insurance or a vehicle’s existing connectivity, goes a long way to assuage participants about some of the inevitable privacy concerns.
• And while smartphones have the benefit of near-ubiquity among drivers, there are many challenges and edge cases that make them less-than-ideal as the primary mechanism for road pricing, especially: poor GPS, drivers turning off their phones, and situations with multiple phones in one car.

ClearRoad-Specific

In which cities is ClearRoad already operating? How many participants are included in each case? And what has been the impact on travel behavior?

ClearRoad currently processes trips for Oregon’s OReGO RUC program and performed a similar role for Washington State’s RUC. The more advanced capabilities described in the webinar have been demonstrated in cities such as New York, London, and Portland.

Participation between these programs is in the thousands of vehicles. RUC programs generally have seen small changes in driver behavior, including less driving and, in certain instances, better driving, though behavior change is not their stated objective.

Is ClearRoad’s system a white-label solution?

Our aim is for this to be white-labeled. We’re agnostic to the specific data sources and the billing & reporting systems used, and frankly, aren’t keen to take on the direct customer service responsibilities. We see Stripe as a good model for plugging into more significant solutions. We imagine that road pricing charges could be integrated into other payments, such as car leases or insurance bills.

What would be the investment cost compared with ClearRoad’s solution for the London Congestion Charge zone, based on the total volume of cars?

We aim for 1/10th of the cost, for both implementation and operation, so for London, perhaps investment in the tens of millions of dollar-range rather than hundreds. We see road pricing “transactions” as analogous to credit card transactions. There, processors such as PayPal or Visa charge fees in the low single-digit range.

Technology

Can E-Zpass be used to attain many of these benefits? E-Zpass infrastructure is already in place. Seems like the ClearRoads model requires a lot of new tech and also for people to have new cars, creating an equity issue.

We certainly wouldn’t suggest cities go and rip out existing infrastructure. Still, the toll tag technology used for E-Zpass and similar systems is severely limited by the requirement of one toll tag reader per lane at each entry point. For NYC’s expected implementation, this translates to multiple readers at over 150 different locations, with a price tag north of $500 million.

Singapore is shifting their ERP to GNSS. What is your opinion of this technology? Is it similar to the GPS-based tech you presented?

GPS is the satellite constellation for North America’s GNSS (you may come across “multi-constellation” hardware). Singapore’s system is temporarily on hold with COVID, but we look forward to seeing the deployment.

Singapore’s is an all-in-one approach that includes expensive standardized satellite devices required to be mounted in all vehicles, real-time digital displays, and enforcement capabilities. It has a price tag to match — around $600 million. In contrast, ClearRoad’s approach would leverage the existing in-vehicle hardware as much as possible and otherwise use off-the-shelf $15 plug-in devices.

Communication

How do you create transparency for drivers around price with such a complex system, potentially involving varying and granular prices? How can you change behavior if drivers don’t understand the link between their choices and costs?

Uber is an excellent example of how complex pricing can be simplified for users. When selecting a ride, Uber presents you with the all-in price, wait time, and estimated arrival for several ride options. Users only get the granular prices (time, distance, surge pricing, etc.) after the ride is complete.

The granularity and real-time pricing of Uber likely won’t work for the general public, at least until autonomous vehicles. In the meantime, there are several techniques for providing drivers with sufficient pricing information to act on. One is to announce prices ahead of time, say the night before. Another is to set a high base fee and provide drivers with multiple mechanisms for receiving. Because of loss-aversion, this may be preferable to a low base fee with penalties.

One element from the presentation that bears repeating is that a congestion pricing system of distributed infrastructure allows individual drivers to experience the system differently. Examples I discussed in the webinar included a small business rate with lower rates for the first hour but quickly grows more expense after that, and a transit incentive program for low-income households along low-ridership bus lines.

Privacy/Security

How do data privacy and surveillance concerns — especially in cities with limited capacity or transparency — factor into a road pricing system that rely on smartphones or plug-in GPS devices? Do cities own the collected data from the vehicles? Is the system able to link a person’s movement to vehicle type/registry?

We actually see the storing of vehicle and customer data as a liability and seek only the minimum amount of data necessary to operate a system. We similarly only share data with governments in the aggregate and allow drivers to purge their data at-will.

Our system architecture, which separates vehicle from customer data, allows us to work with anonymous IDs, providing a firewall between data sources, billing companies, and governments. There’s still more to do to harden our privacy & security processes, but we are in the early stages of putting together our data privacy/security principles. The location services company Bluedot sets a fine example.

The reality is that no technology can provide a perfect fix, and much of the privacy/security issue is political — frankly, once you’re having the discussion with a user, you’ve already lost. Our hope is that the public sector is able to gain citizens’ trust.

How do you address privacy concerns of using on-car infrastructure which tracks people everywhere rather than using off-car infrastructure which only tracks people at certain locations?

Not to sound too snarky, but the data providers in question, not to mention a hundred other data sources, are already capturing data regardless of location. The expectation is that once we have certified a data provider (e.g., an automaker) as sufficiently reliable and accurate, then the data provider will only send us geospatially-relevant data.

Implementation

Does this option limit/change the need for EIS? If we wanted more flexibility or an expanded zone do we still need to have a strict zone for the EIS?

Very prescient since the FHWA has, in fact, held up NYC’s congestion pricing. Our solution should (eventually) simplify EIS submissions and, moreover, flips the EIS on its head! Rather than defining boundaries and policies and then performing (months of ) analysis, this enables a city to describe an outcome or goal and state the potential mechanisms for achieving that outcome.

It’s the difference between saying, “we’ve modeled traffic and expect to reduce emissions by 15%,” and “we are setting an emission reduction goal of 15%. We will achieve this reduction by charging separate, variable per-mile fees on TNCs, trucks, and passenger vehicles in a series of cordons centered around Downtown.

Aren’t alternatives a necessity for equitable congestion pricing? Since you can’t instantly provide transportation choices, even if you devote all the congestion pricing revenue to it, how can you best proceed in the US, equitably?

Yes, if a city or region imposes a fee on driving but doesn’t provide alternatives, then the result is reducing people’s mobility & access, a.k.a. freedom, which is definitely not what we’d advise! So, a few ways to respond:

• Bus lines and associated infrastructure are cheap to deploy and scale. Prior to the Congestion Charge rollout, London increased the number of buses within the Charge area 27%
• Unlike with traditional congestion pricing deployments, ClearRoad’s lightweight approach allows the targeting of specific segments — e.g., freight or TNCs — while still leaving the door open for including passenger vehicle segments down the road.
• Finally, current congestion pricing technologies’ high costs require expensive tolls to pay them off, so anything less than a $10 fee is unlikely to be feasible. A lower-cost system leaves much more flexibility to offer discounts & exemptions, say for low-income households, and generally lower tolls for all.

Recommendations for implementing this in Global South Cities with challenges such as weak law enforcement, technology, and the fact that most of the low-income live farther from the central zones?

An important question! For reference, London is the 5th wealthiest metro area globally, so there’s a broader question of: how can you implement congestion pricing or similar traffic/congestion management in the 90+% of cities where traditional approaches don’t pencil out?

For Global South cities specifically, relying on low-cost, off-the-shelf technologies, including non-smart-phones, is certainly feasible, though we don’t have partnerships in that area yet. To reduce human enforcement costs, I might suggest targeting one vehicle segment at a time, e.g., Jeepneys, rather than all vehicles.

Though low-income populations in Global South cities may live farther from city centers, they tend to take transit, drive small, low-power vehicles, or otherwise share rides. There are several ways that these trips can be discounted or exempted, including based on geofencing, vehicle type, or passenger capacity.

Do you have a breakdown of current operating costs? 30% or more of the revenue seems very high and hard for elected officials to get their heads around.

The key cost drivers are installation & maintenance, old & inefficient back-office technologies, toll-by-plate customer billing, and collection costs. All elements of tolling have inflated costs due to toll operators having effective monopolies on their respective tollways.

While there are plenty of sources available showing tolling revenue, there are unfortunately few useful resources for tolling system installation and operation costs.

Policies

With congestion so influenced by economic costs (time=money) and the potentially competing need to provide equitable access, do you anticipate using your platform to actively reward desired commute behavior (shared rides, off-hour, farther-away parking, etc)? Do you think public or private sector would fund that?

Incentives in one form or another (positive or negative) are integral to all road pricing concepts. We are actively looking at pilots focused on supporting positive behavior change and transit integration, both together and separately, and see a lot of room for exploration. The public sector already funds many behavior change policies, such as free parking or EV access to HOT lanes. I’m less bullish on the private sector being able to act at sufficient scale to meaningfully change congestion or other outcomes, though there are certainly many possibilities.

Are there any reasons not to implement a time-based system, instead of charging a flat one-time or daily fee?

Communicating a flat fee is certainly simpler. Along those lines, a saying from the RUC world is that it will be harder to go from no fee to a 1-cent fee than a 1-cent fee to a dollar. That is, the flat fee can be the first step. We understand the politics of this, but believe it is misguided.

London’s original implementation allowed drivers to pay the Congestion Charge on the TfL website at any point during a day when their vehicle entered the Charge zone. This setup makes variable fee collection almost impossible. Londoners now have multiple automatic options for paying the fee, and our experience with RUC tells us that “active” systems, where responsibility is on the driver to report, are unpopular.

Could the system also be used for a space-sensitive or “passenger per vehicle ratio” charging?

Sure! There are many dimensions on which vehicles could be charged differently. See NUMO’s Periodic Table of Mobility for an excellent rundown. TL;DR As long as there’s data readily accessible, it can be used for pricing.

Enforcement

What about “enforcement”? For example, Express Lanes in Los Angeles are experiencing congestion from major excess demand by violators (“free riders”) that pricing does not control.

Estimates for Express/HOT/Managed Lane across the country have violators accounting for around 1/3 of traffic, so yes, a big concern. The two primary enforcement mechanisms — cameras and patrol officers — are expensive and inaccurate. There are several promising technology solutions to validate occupancy that we are exploring. The best option may be deploying multiple technologies at once.

How do you deal with vehicles that aren’t connected? How do you police non-compliance?

License-plate reading cameras are great for catching violators, but not as the primary means of enforcement due to diminishing returns from installing additional cameras. Until connectivity for all vehicles can be assumed, a secondary data source is necessary. Note, in-vehicle or plug-in hardware can provide compliance data of disconnection, tampering, and validation against other vehicle data sources such as odometer.

One way is to design such systems to be opt-in, so that the only way to be entitled to a discounted rate, for example, would be to participate in this fashion. This already occurs with different rates for toll tags, toll-by-plate, and cash.