Azadeh Dinparastdjadid

Azadeh Dinparastdjadid is senior research scientist at the Safety Research and Best Practices team at Waymo, where she explores road user behavior modeling to establish behavioral benchmarks. Azadeh earned her Ph.D. and M.S. in Industrial Engineering from the University of Wisconsin-Madison in 2020, and received her M.S. in Mechanical Engineering from the University at Buffalo in 2016

---

Presentation

Automated driving system behavior evaluation

The behavioral capability of the automated driving system (ADS) is an important attribute of ADS safety. Behavioral capabilities can be categorized into three broad categories: regulatory compliance, conflict avoidance and collision avoidance. In particular, the annexes in ISO 2022 highlighted the need for more detailed scenario-dependent behavioral evaluation around conflict avoidance. Assuming the identification of a set of scenarios a-priori, this points to the need for objective metrics that can evaluate whether the behavior of the ADS meets the scenario-specific behavioral targets. This presentation aims to discuss computational metrics that can be used as behavioral benchmarks for flagging when ADS behavior does not meet expectations. Such metrics can also be applied to discovering novel scenarios and events with behavioral issues from existing data sets.