Building on its testing of self-driving taxis in Singapore, software company nuTonomy will soon begin evaluating its autonomous technology closer to home.
The company said it will deploy a fully automated vehicle on public roads in Boston by the end of the year. A company executive signed a memorandum of understanding with city officials that will permit nuTonomy to operate in a 191-acre industrial park in Boston’s Seaport area.
Launched as a spinoff from nearby Massachusetts Institute of Technology, the pilot project will be something of a homecoming for the company. But it’s not just proximity to its home base that made Boston an appealing proving ground. It’s the city’s lousy weather and traffic-clogged streets.
“Testing on the streets in Boston will expose our software to the reality of driving in the U.S., including winter driving conditions,” company spokesperson Eric Andrus said.
“We’re throwing open the doors for conversations
with [autonomous driving] companies.” – Kris Carter, Mayor’s Office of New Urban Mechanics, Boston
Winter weather is considered one of the chief challenges for autonomous driving systems. To date, only Ford has claimed to have successfully driven its autonomous cars in the snow, and that was at the Mcity testing facility in Ann Arbor, Michigan, not on public roads.
Should nuTonomy conduct those evaluations during the upcoming winter on public roads and do so without incident—two big question marks—it would be an indication the company is on track to deploy self-driving taxis by its stated goal of 2018, years ahead of the timeline other manufacturers have put forth for launching autonomous operations.
“Testing our self-driving cars so near to nuTonomy’s home is the next step toward our ultimate goal: deployment of a safe, efficient, fully autonomous mobility-on-demand transportation service,” company co-founder Karl Iagnemma said. In addition to weather conditions, Iagnemma said, Boston testing will focus on learning local signage and road markings while gaining further understanding of pedestrian, cyclist, and driver behavior in a complex environment.
For starters, nuTonomy will use a single Renault Zoe electric vehicle for its Boston-based testing, but the fleet is expected to grow. In documents filed with the city, the company says it intends to complete at least 1000 miles with its cars and operate for at least 100 hours. Those operations will include testing at various speeds within legal limits, at various times of day, and in both light and heavy traffic. Once the vehicle proves itself, nuTonomy intends to expand the agreement to encompass more areas of the city.
Humans Might Cause Complications
One other notable aspect of the company’s agreement with the city: Boston may require a police officer to be present in the test area. It’s not that city officials fear the car’s actions; it’s humans they’re apparently worried about.
An officer may be required “in the event of any interference with the the vehicle by other motorists or for other safety reasons.” The officer is not required for testing, “unless interference from other road users necessitate this action.”
It’s a curious provision in two respects. One, any attempts to restrict pedestrians or other drivers from the testing area could inadvertently impact results of the testing. On the flip side, a new report from the University of California Santa Cruz suggests that pedestrians and human road users might attempt to “behave with impunity” and bully self-driving cars.
“Testing on the streets in Boston will expose our software
to the reality of driving in the U.S., including winter
driving conditions.” – Eric Andrus, nuTonomy
The site will be the first U.S. location where the nuTonomy will deploy its autonomous technology, but it is already up and running in Singapore. The company has a “handful” of vehicles operating in Singapore, according to Andrus, but that fleet is expected to grow to a dozen by the end of the year. Depending on the vehicle, the company’s test fleet uses between four and six lidar sensors per car, as well as front- and rear-facing radar and front-mounted cameras that monitor both the path ahead and the cabin of the vehicle.
In August, nuTonomy became the first to reach the milestone of launching an automated taxi service that picks up passengers, beating Uber’s Pittsburgh project to the title by a matter of weeks. In all cases, the projects still require a human behind the wheel.
Boston Becomes Hub for Research
Massachusetts has increasingly sought to attract projects in the advanced-transportation and autonomous-testing realms. In October, the state’s governor, Charlie Baker, issued an executive order that established an autonomous-vehicle working group that will set policy and pave the way for self-driving deployments.
“We’re throwing open the doors for conversations with those companies,” said Kris Carter, co-chair of the Mayor’s Office of New Urban Mechanics in Boston. “From the city side, this process sets some overarching goals for autonomous vehicles, and helping to solve the problems we’re hearing from residents: making the roads safer, making transportation more reliable, and providing access to areas that might not currently be served best.”
In neighboring Somerville, Audi has developed self-parking technology and, through its Urban Future Initiative, has outfitted parking garages for an autonomous era and conducted experiments with advanced traffic-light technology.
In much the same way that Stanford University and Carnegie Mellon University have served as springboards for autonomous development in Silicon Valley and Pittsburgh, respectively, MIT has recently done the same for the greater Boston area. In addition to nuTonomy, another MIT spinoff, Optimus Ride, has started work on Level 4 autonomous technology. Last year, Toyota opened a joint laboratory with MIT that is researching artificial intelligence and autonomous driving.