Tool to estimate the safety impact of vehicle levels of automation on Minnesota roads

Principal Investigator(s):

Gary Davis, Professor, Civil, Environmental and Geo-Engineering

Co-Investigators:

  • John Hourdos, Former Research Associate Professor, Civil, Environmental and Geo-Engineering
  • Raphael Stern, Assistant Professor, Civil, Environmental and Geo-Engineering

Project summary:

Vehicle automation is attracting great interest both commercially and research-wise, and it is expected to disrupt transportation in the years to come. Currently, an increasing number of commercially available vehicles offer Advanced Driver-Assistance Systems (ADAS). Even if commercially available self-driving vehicles won't be available for several decades, the safety, mobility, and environmental benefits envisioned can be achieved with technological advancements already available to the public. MnDOT and other state DOTs as well as the federal government have supported research investigating the operational and environmental effects of adoption rates of vehicles with various levels of automation. Recent findings caution that the resulting future may not be as rosy as once thought. Unfortunately, very little research has focused on the net effect of the biggest CAV selling point: the potential benefits from a reduction in crashes and general increase in road safety. Although this project is, by necessity, performing a thorough investigation of the different flavors of vehicle automation already commercially available as well as exploring their near-future improvements, the main focus is on developing a planning tool to quantify the statewide net safety effect of policies, market, and technology forces affecting the proliferation and actual use of individual ADAS. The proposed tool will use records of actual crashes in Minnesota combined with probabilistic models, relating facts such as vehicle model and age, driver age, and other demographic information with the potential of owning and having activated a specific combination of ADAS features--as well as, given the prevailing road, traffic, and environmental conditions, the probability of this particular ADAS changing the outcome of the event.

Project details: