Will NHTSA rule on automatic braking lead to fewer crashes?

WASHINGTON — An upcoming proposed rule from NHTSA could drive improvements in a widely deployed crash-avoidance technology that has been reliable for many motorists but problematic for some.

The agency as early as this month is expected to issue a proposal to require automatic emergency braking, including pedestrian detection, on all new light-duty vehicles and set minimum performance standards, as mandated in the infrastructure law passed by Congress last year.

The technology has the potential to save lives and reduce injuries by stopping or slowing a vehicle to prevent a collision. But as the capability becomes common across all makes and models — not just luxury vehicles and higher trim levels — more consumer complaints of braking malfunctions have emerged.

Just in February, NHTSA opened two separate investigations involving Tesla and Honda vehicles after receiving hundreds of complaints alleging unexpected activation of the AEB system, characterized by some Tesla owners as "phantom braking."

With its rule-making, NHTSA has an opportunity to address concerns related to occurrences of unintended braking and set the basis for better-performing technologies to develop, safety advocates and tech experts told Automotive News last month.

"Automatic emergency braking saves lives, and it has probably the greatest lifesaving potential of any system that's out there in wide use today," said William Wallace, associate director of safety policy at Consumer Reports.

"Binding performance standards from NHTSA for new and emerging technologies is critical," he added, "and NHTSA needs to be flexible enough to update those standards as the technology matures."

At least 12 automakers, including Honda and Tesla, already have met a 2016 voluntary commitment brokered by NHTSA and the Insurance Institute for Highway Safety to equip at least 95 percent of their light-duty cars and trucks with AEB. The commitment calls for automakers to meet the benchmark for models manufactured between Sept. 1, 2022, and Aug. 31, 2023.

The insurer-funded institute said it expects the voluntary commitment to prevent 42,000 crashes and 20,000 injuries by 2025.

The Alliance for Automotive Innovation, which represents 19 of the 20 participating automakers, said the commitment was a joint effort "to speed up the safety benefits of AEB versus a prolonged rule-making process."

Tesla is not a member of the alliance.

"As AEB grows more ubiquitous, the benefits to the motoring public become more pronounced," John Bozzella, CEO of the group, said in a statement.

Despite the voluntary effort, consumer advocates and safety groups say a regulation is still necessary for holding automakers to their pledge and setting the bar for system safety and performance.

"We've seen that car companies can move quickly" and equip more of their vehicles with the technology, Wallace said, but "it's important for NHTSA to go ahead and get a mandatory rule in place, get binding performance standards in place that apply across the whole market."

Cathy Chase, president of Advocates for Highway and Auto Safety, stressed the word "minimum" and said automakers could even exceed the standards.

"It's really a basement for what automakers must make, which says, 'This system is going to detect, alert and respond if there is an upcoming collision,' " she said. "Anything beyond that is to the benefit of consumers."

As AEB is equipped on more vehicles, complaints concerning the technology have unsurprisingly gone up, said Michael Brooks, acting executive director at the Center for Auto Safety.

Starting in 2015 with a now-closed safety probe into certain Jeep Grand Cherokees over complaints alleging unexpected braking with no imminent collision, the center has tracked three other NHTSA investigations directly involving unintended activations of AEB, plus a fatal highway crash involving a 2015 Tesla Model S operating on Autopilot where the AEB system did not activate.

Two of the more recent investigations cover an estimated 416,000 2021-22 Tesla Model 3 and Model Y electric vehicles equipped with Autopilot and 1.7 million 2018-19 Honda Accord and 2017-19 Honda CR-V vehicles.

The center, too, continues to urge Nissan to recall 2017-18 Rogue and Rogue Sport vehicles, which are part of a still-open investigation from 2019 that has logged more than 1,400 complaints and field reports of braking problems associated with the AEB system and at least five injuries and 14 crashes.

"It could be the software. It could be the sensors. It could be the cameras — a lot of things depend on the type of system that's in use," Brooks said. "But whatever the case may be, there needs to be something in the standard to prevent these false activations."

Tesla Model 3, Model S, Model X and Model Y vehicles built for the North American market are no longer equipped with radar and, instead, rely on cameras for AEB and other active safety features, according to the EV maker.

The Honda models included in NHTSA's safety probe are equipped with "a version of Honda's Collision Mitigation Braking System that features both cameras and radar," Honda spokesman Chris Martin said in an email. Honda has not yet used lidar on any vehicle sold in the U.S., he noted.

The vast majority of AEB systems are camera-based, according to Sam Abuelsamid, a principal analyst at consulting firm Guidehouse Insights.

Though cost-effective, camera-only-based systems "don't work at night," he said.

"They don't work at even moderate speeds above 25 miles an hour or so, and they just can't be relied on."

While cost per vehicle depends on the hardware suite used, more advanced systems with lidar could be done for less than $200, including camera and basic radar, Abuelsamid said. But NHTSA's regulation should be technology-agnostic and focus on outcome rather than specifying a certain approach to achieving that outcome, he stressed.

"It should not necessarily specify that you have to use radar or lidar along with a camera or any other specific type of sensor," Abuelsamid said. "If a manufacturer is able to prove that they can do it with a pair of near-infrared cameras and no other sensors, that's fine. It's all about the outcome, not about how you get that outcome."

While the investigations allow NHTSA to examine potential defects and determine whether the braking malfunctions are rare and predictable, the upcoming proposed rule-making will specify a test procedure for AEB, effectively measuring automakers' compliance with minimum performance standards and possibly tackling situations where the systems activate when there is no imminent crash.

"If the engineering investigations uncover some common sources of false activations, then I think what you might see … is some test procedures built around those common sources to make sure that at least the AEB systems that are mandated are immune to those sources," said David Zuby, chief research officer at IIHS.

Bridge expansion joints, railroad tracks, metal trench plates, smoke and steam are widely known as potential triggers for false-positive activations, he noted.

The institute, for its part, plans to release a new nighttime test and rating system for AEB systems this year, following a study that examined their lack of effectiveness on dark roads.

"Ultimately, technology solves this — better technology, better sensors, better software," Zuby said.

"In terms of writing a regulation, it will come down to striking the right balance between effectiveness at preventing crashes without inadvertently creating too much sensitivity to situations that shouldn't create activations."