Thomas Breen photo
City Plan’s Leslie Radcliffe: It will take “one failure to be a tragedy.”
Stantec
Driverless shuttles: Coming soon to a hospital near you?
A plan to test driverless shuttles on New Haven streets advanced Wednesday night — with dissenters raising fears about public safety and the loss of human drivers’ jobs.
That conversation and vote took place Wednesday night during the regular monthly City Plan Commission meeting in the basement meeting room of the 200 Orange St. municipal office building.
The commissioners voted 3 – 2 in support of recommending that the city apply to be a part of the state’s Fully Autonomous Vehicle Testing Pilot Program.
Stantec Principal Craig Lewis and city transit chief Doug Hausladen: This pilot would give New Haven a “seat at the table.”
The item now advances to the Board of Alders for a committee public hearing and a subsequent final vote — not on whether or not the city should start a driverless car program, but on whether or not it should apply to become one of four Connecticut municipalities eligible to receive state support to build out such a program sometime down the road.
City Department of Transportation, Traffic & Parking Director Doug Hausladen joined Craig Lewis of the Stamford-based consultancy Stantec to give the commissioners an updated version of a pitch he made to the city’s Traffic Authority last April about this very same pilot program application.
“This is not an approval to launch a driverless transit system,” Hausladen stressed Wednesday.
Rather, he was asking for an approval for the city to become one of four municipalities in the state where this emerging transportation technology can be tested in a thoughtful, planned way.
As part of the city’s application to participate in this state-backed pilot program, Lewis said, Stantec and the city have put together a few preliminary recommendations on what a potential driverless vehicle pilot could look like in New Haven.
He said the vehicles Stantec would recommend are squat, 16-foot-long driverless shuttles that can hold eight to 16 passengers at a time.
They do not have a driver’s wheel, and would not have a driver. Instead, they would use fiber optic technology to interact with signals and traffic, fixed landmarks and moving pedestrians and cyclists and other vehicles.
The driverless shuttles would go no faster than 10 to 15 miles per hour, and would be able to run for three to 10 hours per battery charge.
Lewis said there are somewhere between seven and 16 authorized manufacturers of this technology in the marketplace today, and that there have been 70 different pilot programs featuring these types of slow-moving autonomous shuttles — including pilots in San Francisco, Columbus, Las Vegas, Detroit, and Pittsburgh.
“There have been zero accidents,” he said. Some autonomous shuttles have been hit by “drivered” cars, and one higher-speed driverless car in Arizona did kill a pedestrian in 2018.
The immediate goal of such a pilot program would be to better understand how autonomous vehicles work, Lewis said, with the longer-term goal of providing “faster, less expensive, and more efficient transportation services.”
Stantec
Lewis and Hausladen said that the two initial routes that the city plans to recommend for a potential autonomous vehicle pilot program would overlap with two existing roughly 1.7‑mile Yale New Haven Hospital shuttle routes that connect the St. Raphael Campus and the York Street campus.
The vehicles would travel along George Street, Park Street, Legion Avenue, and Sherman Avenue on weekdays between 8 a.m. and 6 p.m. One of those routes has 14 traffic signals and three pedestrian crossings, the other 11 signals and four pedestrian crossings.
As of right now, federal law requires that driverless vehicles like these have a human on board responsible for providing some kind of emergency backup. During the pilot, those staffers would double as “greeters,” Lewis said, and would likely be phased out as the technology progresses.
Hausladen said the hospital shuttle routes were selected because the hospital is the largest employer in the state, meaning that there are plenty of potential riders in that relatively small area of the city looking to get from one spot to another. Mapping the shuttle pilot onto existing routes would also allow the city and the hospital to directly compare the performances of an autonomous vehicle with the existing “drivered” shuttles.
He said the other three applicants for this state pilot eligibility program are UConn, Stamford, and Bradley Airport. He said the state Office of Policy and Management (OPM) has not yet announced how much money will be made available to applicants allowed to participate in such a program.
Participating in this pilot would mean that New Haven transportation and public safety and emergency response officials would have a “seat at the table” in determining how this driverless vehicle technology will be rolled out and regulated across Connecticut in the years and decades to come, Hausladen argued.
Commission Chair Ed Mattison.
City Plan Commission Chair Ed Mattison agreed. “It does feel that doing this in a planned way” is the smartest way to proceed, he said.
If the city decides not to apply or participate in such a pilot program, he warned, then private operators will likely test out their own technologies on their own dimes in New Haven. “Before it just happens,” he said, the city and public sector should investigate and figure out the best regulations.
Commissioner Adam Marchand (center).
“I agree that it’s better to have a planful approach,” said Westville Alder and City Plan Commissioner Adam Marchand.
He also warned that Legion Avenue is not a “slow-speed” stretch of road by any means, and said he worries about the speed differentials between these slow-moving shuttles and surrounding traffic.
“I Have Grave Concerns”
Attendees at Wednesday night’s City Plan Commission meeting.
After Lewis and Hausladen finished their presentation and opened the floor to questions, Commission Vice-Chair Leslie Radcliffe quickly emerged as the most skeptical of the group.
Her concerns centered around the safety, and wisdom, of testing out such an experimental new transportation technology in one of the most densely traversed areas of the city — filled not just with cars, but pedestrians, cyclists, and infirm people trying to get to the hospital.
“It seems to be driven mostly by economics and not by safety,” Radcliffe said upon hearing from Lewis and Hausladen that lower expenses and greater efficiency are two of the big goals of a potential pilot program.
The motion-sensor technology may be able to discern between fixed street infrastructure and curves in the road and moving cars and crossing pedestrians, but what about if — or when — one vehicle makes one mistake?
“It’s only going to take one failure for it to be a tragedy,” she said.
She said she feels uncomfortable at the prospect of placing city residents and visitors in the role of “involuntary participants” of a driverless car tech test.
Why can’t this be done in a more remote and less traveled area of the city? she asked. Or, even better, in a secluded track designed for just such testing?
“I have grave concerns.”
Hausladen said that the risks and the rewards of piloting such a program in such a highly-trafficked area of the city go hand in hand. There’s a clear need for shuttle transport between the two hospitals, with plenty of passengers standing to benefit from the implementation of more reliable transit options. On the flip side, there are always plenty of unknowns whenever one tests new technology.
“Right now, I really have questions,” he admitted. “I don’t have answers.” Thus the city’s interest in applying to be eligible to participate in a pilot — and not to launch a pilot right away.
Commissioner Ernest Pagan: What about human drivers’ jobs?
Commissioner Ernest Pagan added another set of concerns — around employment — to the conversation.
This project sounds like it will lead to the unemployment of lots of bus drivers, he said. And those jobs lost will only be made up for in part by lesser-paid “greeter” positions.
Hausladen recognized that that is the two-edged sword of automation. More efficient and reliable transportation options would be a boon for passengers, he said, but could lead to the short-term loss of certain types of work.
“We have to be worried about the cost of delivering service,” he said, “and the opportunities we miss because we couldn’t deliver more service.”
Ultimately, Radcliffe and Pagan voted against recommending that the alders approve the proposed order, while Mattison, Marchand, and Commissioner Kevin D’Adamo voted in support.
What makes Hausladen think that 15 mph dumb buses will be "more efficient and reliable" than adaptive human drivers?