Remarks as prepared for delivery.
Good morning and welcome to the National Transportation Safety Board (NTSB). I’m Jennifer Homendy and I’m honored to serve as Chair of the NTSB.
With me today are my Board colleagues. But before I introduce them, a big congratulations to Mike Graham, who has been appointed as Vice Chairman of the NTSB!
Mike has been on the Board since January 2020. Since then, he’s advocated for connected vehicle technology and ending distracted driving on our roads. A pilot, he’s been vocal in advancing our recommendations on flight data monitoring and promoting the use of safety management systems (SMS) in all modes of transportation.
Congratulations, Vice Chairman Graham!
Moving on, I now want to introduce Member Tom Chapman and our newest Board Member: John DeLeeuw.
John comes to us from American Airlines, where he most recently served as Managing Director of Safety and Efficiency. John is an experienced pilot with more than four decades flying for the United States Air Force and as a captain at American Airlines, most recently flying the 787.
Capt. DeLeeuw’s experience in establishing and supporting safety culture, managing complex investigations, and communicating about safety will be an asset to our agency and to the traveling public — as will his experience serving as a party coordinator in NTSB investigations. Welcome, Member DeLeeuw!
Today’s meeting is open to the public, in accordance with the Government in the Sunshine Act.
We’re here to discuss the final investigation report into two fatal, rear-end collisions. In both cases, the striking vehicle was a 2022 Ford Mach-E operating in hands-free partial driving automation mode.
The first crash occurred on February 24, 2024, in San Antonio, Texas, where the Ford driver rear-ended a stationary Honda CR-V in the center lane of eastbound Interstate 10.
The second crash occurred eight days later, on March 3, 2024, in Philadelphia, Pennsylvania. The Ford driver rear-ended a stationary Hyundai Elantra in the left lane of northbound Interstate 95, triggering a sequence of crashes involving two other vehicles.
Tragically, in these two crashes combined, three people died and two people sustained minor physical injuries.
I say “physical” because the trauma that occurs from any crash can be life long. I know this from my own experience of being rear-ended at high speed while stationary at a red light in 2021.
To the families and friends who lost loved ones in these crashes, as well as the survivors who were injured, on behalf of the entire agency, I want to extend my deepest sympathies.
Please know that our work doesn’t end with the issuance of our final investigation report. After today’s proceedings, we’ll work diligently to implement our safety recommendations so that no other family has to endure such tremendous loss.
The NTSB believes that the only acceptable number of lives lost on our nation’s roads … or in any mode of transportation … is zero.
A lot of people believe that zero is unachievable. To them I’d ask: What if it were your family?
We are far from zero.
In 2024, according to early estimates from the National Highway Traffic Safety Administration, there were 39,345 deaths on our nation’s roads — down from more than 43,000 in 2021.
While certainly an improvement, nearly 40,000 deaths is nothing to celebrate. That translates to about 110 people dying every single day.
Millions are injured in crashes each year. The three biggest contributors are impaired driving, speeding, and distraction — all of which we’ll talk about in relation to these crashes.
I often hear from manufacturers that automation in vehicles is the solution to what is clearly a public health crisis on our roads. I think they need to be very careful with these claims.
I support myriad safety technologies; after all, “safer vehicles” is part of a Safe System Approach, which this Board has endorsed.
In my own vehicle, which is a SAE Level 2, partially automated vehicle, I often use Subaru’s “EyeSight” adaptive cruise control. But I steer clear of lane-centering because I know my own limitations: I will get distracted.
Certainly, I look forward to the day when we can say we’ve eliminated impaired driving, speeding, and distraction. And I’m hopeful in-vehicle technologies can help make that a reality.
But we have a long way to go.
Two technologies the Board has long advocated for are forward collision warning and automatic emergency braking systems.
A 2015 NTSB Special Investigation Report concluded that, when combined, these systems “could significantly reduce the frequency and severity of rear-end crashes.”
This was confirmed two years later by the Insurance Institute for Highway Safety, which found that rear-end crashes dropped 50% when vehicles were equipped with both systems.
The study estimated that widespread use of these technologies could have prevented nearly 1 million crashes and over 400,000 injuries in the U.S. in one year alone.
The evidence for partial automation systems is not as promising.
We are fast approaching the 10-year anniversary of the first fatal partial automation crash in the U.S., which occurred on May 7, 2016, in Williston, Florida.
The NTSB has been concerned about the safe implementation of these systems ever since.
More tragedies were to follow:
- January 2018; Culver City, California.
- March 2018; Mountainview, California, where one person died and another was injured.
- March 2019; Delray Beach, Florida, where one person died in a sequence that closely resembles the 2016 Williston crash.
Overreliance on automation, which can lead to complacency, played a role in all these tragedies.
Decades of research across various industries, including aviation, concludes that humans are poorly suited for automation monitoring tasks and susceptible to complacency.
A 2001 study revealed that air traffic controllers are less effective at detecting conflicts between aircraft when they assume a passive, monitoring role, as compared to actively controlling the airspace. The effect is even more pronounced under high traffic loads.
In 2024, the Insurance Institute for Highway Safety (IIHS) demonstrated that the presence of partial automation systems did not result in lower crash rates and concluded these systems function primarily as convenience features rather than safety enhancements.
That same year, a joint study by the IIHS and MIT Center for Transportation and Logistics studied how drivers adapt to partial automation safeguards over time. The results are extremely concerning: Drivers adjust, learning to internalize safeguard sequences and find windows of opportunity to do non-driving-related activities.
In fact, numerous studies have found that, when operating a vehicle partial automation mode, drivers are more likely to engage in non-driving-related activities than if they were manually driving.
You will soon hear in detail how insufficient federal oversight of partial driving automation systems is leading to catastrophic outcomes.
In a perfect world, the regulator would set minimum performance standards for safety technologies, test them, rate them, and put that information somewhere all auto-buyers will find it: On the car window sticker.
Just like crashworthiness ratings. Just like the rest of the world does.
But the U.S. government doesn’t do this. Automakers are left to create and test to their own standards.
In the year 2026, it’s simply unacceptable that a car in the U.S. can achieve a five-star safety rating even if it has ZERO collision-prevention and other driver assistance technologies.
Our European counterparts are setting standards on what safety technologies must be on vehicles and how they must perform and issuing regulations to prevent unintended consequences from mandating such technologies; they are certainly well ahead of us when it comes to safety ratings.
Thankfully, organizations like IIHS and Consumer Reports have stepped up to fill the void. Their tests reveal that not all manufacturers’ safety technologies are created equal.
This is a fact many U.S. consumers are unaware of.
Let’s now turn to today’s meeting.
Each Board Member has studied the draft report and met individually with the investigative team, but this is the first time we’ve gathered as a Board to discuss the report.
The public docket for both crash investigations was released on June 25, 2025, and contains hundreds of pages of additional relevant material. Both dockets are available on our website at NTSB.gov.
Before I turn things over to our Deputy Managing Director, Brian Curtis, I’d like to thank the agencies that responded to these tragedies.
Effective post-crash care is central to our ultimate goal of zero roadway fatalities. As you will hear shortly, emergency responders arrived at both crash scenes within minutes.
I’d also like to thank Honda for assisting us in San Antonio with the post-crash inspection of the CR-V, and the National Highway Traffic Safety Administration for notifying us of the crash.
And finally, I want to acknowledge our colleagues in the Office of Highway Safety and throughout the NTSB for their hard work on this investigation and for their excellent work in putting together this draft report, proposed findings, probable causes, and safety recommendations. You all did a phenomenal job.
Mr. Curtis, you’re now recognized.
