Project: #12 Prediction and Behaviors for Driver Assistance and Socially Cooperative Autonomous Driving Progress Report - Reporting Period Ending: Sept. 30, 2018 Principal Investigator: John Dolan Status: Completed Start Date: Jan. 1, 2017 End Date: Aug. 31, 2018 Research Type: None Grant Type: Technology Transfer Grant Program: FAST Act - Mobility National (2016 - 2022) Grant Cycle: 2017 Mobility21 UTC Progress Report (Last Updated: Sept. 28, 2018, 8:25 a.m.) % Project Completed to Date: 100 % Grant Award Expended: 100 % Match Expended & Document: 100 USDOT Requirements Accomplishments In earlier work, we developed a Probabilistic Graphical Model (PGM)-based method for estimating the yield/not yield intention of a car merging onto a main road from an entrance ramp. During the most recent reporting period, we eliminated two assumptions in the previous model: 1) a fixed merging point for all merging agents, which is hard to obtain before the merging vehicles make the lane change; 2) Perfect velocity measurement, which requires sophisticated perception systems. We validated the performance of our method both on real merging data and using a designed merging strategy in simulation, and show significant improvements compared with previous methods. Parameter design is also discussed by experiments. The new method is computationally efficient, and robust against sensing uncertainty. Impacts The primary impact of the project's work is improved ability of an autonomous car to deal safely with merging traffic in highway or urban situations, especially in situations in which the on-road car is following a leading car and the merge ramp has multiple cars attempting to merge. We also performed initial work on a POMDP (Partially Observable Markov Decision Process)-based approach to autonomous driving behaviors at traffic circles, or roundabouts. Other We have collected and analyzed human driving data on a testbed based on a Logitech steering wheel and pedal system coupled with both the VTD and CARLA driving simulators in order to collect human driving data in various scenarios. Outcomes New Partners None Issues None