The Evolution of Vehicle Safety: The Intersection of Automation and Human Behavior

Oct 25, 2017

Session Description

Technological innovation in transportation automation over the coming decades offers a promise to disrupt how we live and move. The ability of society to adapt to the pace of change is being increasingly challenged. How will we safely transition drivers, and other road users, into a mixed fleet environment?  How can urban planners and policymakers accelerate the local benefits of such technology?  What can lessons from the past tell us about the need for consumer education, public policy, and a human centric design philosophy?



Moderator: Scott MillerGeneral Manager, Collaborative Safety Research Center, Toyota Motor North America

Bryan ReimerResearch Scientist, MIT AgeLab & Associate Director, New England University Transportation Center, MIT

Jeffrey Michael, Associate Administrator, Research and Program Development, National Highway Traffic Safety Administration


Leave a question or comment here to start a discussion. Click "Reply" to respond to a specific comment. All comments are public.



  1. Tesla has offered Autopilot for over a year now, without any regard to federal or state legislation. What is your opinion on their approach to automated transportation? Is the Tesla Autopilot reliable enough to be considered “safe”? How can we help legislators keep up with the rapid pace of change in this industry?

    • Matthew, they jury on the safety of Autopilot is still out. One could based upon the data at this point argue strongly for either side. While there are risks with driving (at any point with any set of technology) the complex aspect of assessing safety have a lot to do with the selection of a proper baseline for risk (another topic of open debate). For instance, is Autopilot safer than manual driving on the highway, safer that driving with adaptive cruise control? Alternatively, does one include operation outside of the design domain (where the recent NTSB report critiqued the design) directly into the equation or look at it separately? In any case, only Tesla likely has the data at this point to know for certain. I am leading an effort at MIT where we have collected over 250,000 miles of data from Autopilot equipped vehicles (telemetry, video from inside and around the car) and are looking deeply at use characteristics of the system. What I will share is that among the special population of users (high socioeconomic, lead adopters etc.) use of Autopilot is greater than our expectations. Data on transfers of control and more traditional correlates of risk (acceleration triggers etc.) forthcoming. There is a large need for more independently funded research investment in understanding Autopilot and a range of related technologies. We hope to release a paper or two later this year that begin to provide some information the world needs to better consider the positive and negative aspects of this one game changing technology.

  2. There is a lot of discussion on how autonomous vehicles and smart mobility systems can serve people who currently do not have the ability to operate vehicles today including older adults and people living with disability. What is the role that people trained and experienced in working with these communities can play in developing these smart mobility systems? How do we implement these technologies in a less siloed way?

  3. What do you think of Volvo’s incremental approach, going from adaptive cruise control, to city safety auto stop, to pedestrian and animal recognition, to run off mitigation, driver assist, and partnering with Uber in Pittsburgh on porto-autonomous vehicles? They seem like a leader in this area who doesn’t get much attention in America.

  4. 2 concerns looking at infrastructure history: will the increasing sophistication of these vehicles reduce access, either because of cost or user information limits? Will there be infrastructure requirements that cause localities to displace needed investments in our vast existing infrastructure?

  5. Bryan’s point that public policy is needed to slow technology a bit and get it aligned with the common good is paramount for all of us in the audience.

  6. What I don’t like about automation is that it is often auto-centric. We need to get away from the isolation and traffic congestion of solo driving in cars, automated or not. We need to support public transit instead, and more of it is becoming automated. Elevated light rail can be solar-powered and safe as in Cybertran. Check it out.


Submit a Comment

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Share This