Reliable Deep Learning for Data-Driven Mobility Prediction under Uncertainty for Off-Road Autonomous Ground Vehicles

Project Team

Principal Investigator

Zissimos Mourelatos, Oakland University Zhen Hu, University of Michigan, Dearborn

Government

David Gorsich, Amandeep Singh, U.S. Army GVSC

Industry

Guga Gugaratshan, HBM nCode Federal LLC

Yan Fu, Ford Motor Company

Student

Mohammed Billal Kamal, Dakota Barthlow, Oakland University

Yixuan Liu, U. of Michigan - Dearborn

Project Summary

Project #1.A90 began Jan. 2021 and was completed Dec. 2023.

Autonomous assets and connected vehicles allow the collection of large-volumes of vehicle mobility data under various terrain conditions. While AI has the potential of developing a powerful data-driven mobility predictive model, the accuracy of an AI model can be severely affected by the data quality used for training. Reliability of AI models is critical for mission-critical applications.

The huge amount of information, however, has not yet been explored in developing the Next Generation NATO Reference Mobility Model (NG-NRMM). Leveraging “big mobility data” has great potential to accelerate the development of a NG-NRMM (1) by eliminating simplifications and assumptions in simulations, and (2) by accounting for high-dimensional environmental (e.g. soil) and vehicle system parameters in a holistic manner.

The overall objective of this research is to develop easy-to-use and scalable AI-based data-driven mobility predictive models for off-road autonomous ground vehicles, with quantified prediction of uncertainty and assured prediction reliability. To achieve this objective, the following three fundamental research questions (RQ) are addressed using a systematic deep learning-based mobility prediction framework:

RQ1: How to build an off-road mobility predictive model using big data analytics and AI?

RQ2: How to quantify various sources of uncertainty in the AI predictive models?

RQ3: How to collect data for effective V&V of AI-based mobility prediction models?

1.A90

Publications:

Liu, Y., Jiang, C., Zhang, X., Mourelatos, Z. P., Barthlow, D., Gorsich, D., … & Hu, Z. (2022). Reliability-based multivehicle path planning under uncertainty using a bio-inspired approach. Journal of Mechanical Design, 144(9), 091701.
Liu, Y., Barthlow, D., Mourelatos, Z. P., Zeng, J., Gorsich, D., Singh, A., & Hu, Z. (2022). Mobility prediction of off-road ground vehicles using a dynamic ensemble of NARX models. Journal of Mechanical Design, 144(9), 091709.
Yin, J., Hu, Z., Mourelatos, Z. P., Gorsich, D., Singh, A., & Tau, S. (2023). Efficient reliability-based path planning of off-road autonomous ground vehicles through the coupling of surrogate modeling and RRT. IEEE Transactions on Intelligent Transportation Systems.
Yin, J., Li, L., Mourelatos, Z. P., Liu, Y., Gorsich, D., Singh, A., … & Hu, Z. (2023). Reliable global path planning of off-road autonomous ground vehicles under uncertain terrain conditions. IEEE Transactions on Intelligent Vehicles.
Yin, J., Hu, Z., Mourelatos, Z. P., Gorsich, D., Singh, A., & Tau, S. (2023, August). An Efficient Surrogate Modeling Method for Reliability-Based Global Path Planning of Off-Road Autonomous Ground Vehicles. In International Design Engineering Technical Conferences and Computers and Information in Engineering Conference (Vol. 87318, p. V03BT03A052). American Society of Mechanical Engineers.
Liu, Y., Jiang, C., Zhang, X., Mourelatos, Z. P., Hu, Z., & Gorsich, D. (2022). Simulation-Based Multi-Vehicle Path Planning Under Uncertainty Using Physarum-Based Approach. In AIAA SCITECH 2022 Forum (p. 1869).