Register now After registration you will be able to apply for this opportunity online.
Submodular optimization for scenario sampling in autonomous vehicles safety testing
A key barrier hindering the swift introduction of autonomous vehicles (AVs) in real-world contexts is the challenge in establishing clear safety benchmarks. Specifically, the issue of systematically assessing both performance and safety remains a significant stumbling block within the industry.
This challenge is mainly twofold: Firstly, how can we identify an ideal scenario set to evaluate the vehicle's performance within a targeted Operational Design Domain (ODD) and what criteria would be useful in amplifying or paring down this set?
Secondly, how do we determine a substantial stopping criteria for the evaluation campaign, and what level of confidence should be attached to the observed performances?
Not specified
Tasks
1. Get familiar with the relevant literature (mainly Waymo’s safety case approach, Motional’s paper on scenario sampling, counterfactual safety measures)
2. Devise and validate a principeled mathematical model for scenario-based testing to infer the safety of an AV. This includes both the development of mathematical/statistical model as well as its experimental validation on real data (Commonroad scenarios, Waymo Open Motion dataset, Nuplan,…).
3. Wrap-up results in thesis report and/or publications
Expected Outcome
The student is expected to show a good working ethic and the ability to collaborate well with other people. From a scientific point of view, for a SP we expect the student to be able to implement or iterate over only a subset of the tasks. We do believe that the product of this research can be published in top robotics conferences/journal, thus a strong interest in research and academic writing is also required.
Tasks 1. Get familiar with the relevant literature (mainly Waymo’s safety case approach, Motional’s paper on scenario sampling, counterfactual safety measures) 2. Devise and validate a principeled mathematical model for scenario-based testing to infer the safety of an AV. This includes both the development of mathematical/statistical model as well as its experimental validation on real data (Commonroad scenarios, Waymo Open Motion dataset, Nuplan,…). 3. Wrap-up results in thesis report and/or publications
Expected Outcome The student is expected to show a good working ethic and the ability to collaborate well with other people. From a scientific point of view, for a SP we expect the student to be able to implement or iterate over only a subset of the tasks. We do believe that the product of this research can be published in top robotics conferences/journal, thus a strong interest in research and academic writing is also required.