April Updates on AgeLab Driving Research
by Niels Wu
April is National Distracted Driving Awareness Month, which highlights efforts to reduce traffic fatalities caused by distracted driving. As vehicle technologies evolve, they create new opportunities to support drivers, while also introducing new ways for attention to be divided away from the driving task.
The MIT AgeLab’s AI With Awareness for Real-World Engagements (AWARE) Initiative seeks to address distracted driving by informing the design of systems that improve driver safety and experience rather than compete with the driving task. As part of this work, AgeLab Research Scientist Pnina Gershon collaborated with Alex Mueller from the Insurance Institute for Highway Safety (IIHS) on a joint study on how drivers adapt to partial automation by finding “windows of opportunity” to engage in non-driving tasks.
The study was referenced by National Transportation Safety Board (NTSB) Chairwoman Jennifer Homendy during the NTSB Board meeting on the investigation (mention at 10m 35s) into two crashes involving Ford’s hands-free driving technology, BlueCruise, in San Antonio and Philadelphia. Both crashes involved driver distraction. In one crash, driver interaction with the infotainment system was identified as a contributing factor. In the other, phone use and alcohol impairment. These sources of distraction are all research priorities under the AWARE Initiative.
This month, AgeLab postdoctoral researcher Bridge Zhao also presented a paper authored by himself and Gershon, “From Prediction to Design: Using Context-Aware Graph Neural Networks and Explainable AI to Anticipate Transfer-of-Control,” at the 2026 World Congress Experience. Leveraging MIT naturalistic driving data, the paper explores whether it is possible to predict when a driver will hand control over to automation, or when they will take it back.
Findings show that the transfer of control of the vehicle does not happen suddenly. Instead, it is associated with measurable changes in drivers’ behavior, vehicle dynamics, and the overall context of driving. These changes were detectable several seconds ahead of the transfer, a fact that driver-assistance systems could leverage to anticipate changes in vehicle control and thus improve driver safety.
Drive safe, this month and every month!
- Share
