Skip to main content

Trust but verify: The challenge of carpool lanes

Published Date
Author Joseph Averkamp

Transportation agencies are always seeking approaches to increase throughput on existing roadways. The truth is that we can’t build enough infrastructure to accommodate all of the transportation that the public wants to consume.

“We can’t build enough infrastructure to accommodate all of the transportation that the public wants to consume.” — Joseph Averkamp, Senior Director, Technology, Policy and Technical strategy

“Managed Lanes” are one tool in the transportation agency’s tool box. They are typically referred to as High Occupancy Vehicle (HOV) lanes or High Occupancy Toll (HOT) lanes. These lanes can require that the vehicle have multiple occupants (HOV2+, HOV3+). They may also require that a Single Occupant Vehicle (SOV) pay a toll. In this manner, the agency creates incentives for the travelling public to ride together, and increases throughput of people on the roadway.

In creating the HOV/HOT solution, as is often the case, a new set of problems is generated. The biggest challenge related to HOV/HOT lanes then becomes how to verify that the vehicles in the HOV/HOT lanes meet the requirements to be in the lane. For the most part, HOV/HOT lanes rely on trust — we count on people being good actors who won’t take advantage of the system. And then to improve the likelihood of compliance we set up enforcement methods for HOV/HOT lanes.

Subscribe to email updates and receive the latest transportation industry trends and tips.

Traditionally, the best way to assure that drivers comply with the requirements of HOV/HOT lanes is to place enforcement officers on the roadside to count occupants. The officers then proceed to pull over and cite violators. This approach creates four major concerns:

  • Equity: You can’t possibly catch all the violators. And then when drivers who are in HOV/HOT-compliant vehicles see violators all around them they develop resentment. People want some assurance that rules are being enforced in an equitable, non-random fashion.

Could “trust but verify” come to a highway near you? http://ctt.ec/iJdXn+ #automation

 
  • Safety: Officers stationed on the roadside and pulling over violator vehicles always present a safety concern. While roadside interdictions will continue to be the primary method for citing most moving violations, any time we can reduce the number of pull overs for violations, we make the roadway safer for the officers and for the driving public. Officers are performing a great public service and, if possible, we should reduce their exposure to vehicles cruising past at 70-80 miles per hour just feet away from them.
  • Congestion: When an enforcement vehicle is placed on the roadside, this alone generates congestion. The driving public will gawk, slow down, and check their speed because they observe an enforcement vehicle. They may be thinking “What are the police doing here?” “Do they have someone pulled over?” “I should probably post this to Waze.” All of these activities cause traffic slowdowns, and if you recall our original purpose, the intent of the HOV/HOT lane is to reduce congestion — the placement of an enforcement vehicle on the roadside can be counterproductive. If we could devise a system that would allow us to handle violations remotely or further down the road at a wider pull over point, we could reduce congestion effects due to enforcement activity.
  • Accuracy/Reliability: Among the challenges for using human eyeballs to identify violators is that vehicles are traveling at high speeds, often in heavy congestion. Some roadways can experience 500 to 1,500 vehicles per lane per hour. This means that a roadside observer must look at 8-24 vehicles every minute and make a snap judgment—“Was that a Single Occupant Vehicle (SOV) or a High Occupancy Vehicle (HOV)?” “Was there someone in the back seat?” This translates to seeing a vehicle every 3-7 seconds. Looking away means missing a potential violator. And then the system becomes random at best — the determinant of whether a driver gets caught is whether the enforcement officer is looking up at the time or not. This is not to say that enforcement officers are deficient in their jobs, but rather that we have given them a very difficult task to perform.

So the challenges for managing an HOV/HOT lane are:

  • The system relies on trust. Drivers must honestly self-declare if they should be in the HOV/HOT lane. How can we trust but verify?
  • People want some evidence of equity. They think “If I follow the rules, shouldn’t others be required to follow the rules as well?” How can we achieve equity? How can we establish a system where enforcement isn’t random?
  • Improve accuracy and reliability. Can we improve the ability of officers to enforce and improve their accuracy? Is accuracy actually a problem for humans on the roadside?
  • Make the HOV/HOT lane serve its original purpose. Can we reduce congestion, improve throughput, and do these tasks while improving safety?

We developed this Vehicle Passenger Detection System to address these problems, and we tested the solution with Caltrans on Interstate 5 in southern California. We were awarded a Best of ITS Award 2016 for this project, which was conducted with Caltrans.

I will share more details from the study in this blog on Monday, July 25.

Print