Naturalistic Data Collection and Analytics for Vehicle and e-Scooter Conflicts

Our goal is to develop a model of the incidents between electric scooters (e-scooters) and vehicles, by collecting and analyzing naturalistic scooter riding data. With this observation model, we can then estimate safety issues e-scooter riders may face on the road and develop necessary precautions.

The Problem 



Over the last several years, major U.S. cities have launched and expanded E-scooter fleets. In fact, there were over 85,000 scooters in service in more than 100 cities around the country by the end of 2018, totaling 38 million-plus trips. With so many scooters on city streets, we must develop an understanding of potential new safety concerns between scooters and vehicles. 

The Question 

In 2018, there were approximately 14,600 e-scooter injury accidents in the U.S., up from just 4,500 in 2014. What potential safety issues exist between e-scooters and cars, while sharing pavement with one another, and what can we do to minimize risks?

What We Did



To propose countermeasures to e-scooter and vehicle accidents, we first need to outline and quantify the necessary safety scenarios. As such, we have based our studies around the following:

 

  1. Development and evaluation of effective e-scooter and vehicle-based data collection systems.
  2. Development of algorithms for accurate scenario reconstruction.
  3. Development of data analysis algorithms to examine naturalistic driving data, including real-world video clips of e-scooter and vehicle crashes.  

The Result 

Research to develop crash models and countermeasures is currently ongoing.  

This project is in collaboration with Indiana University-Purdue University Indianapolis (IUPUI).