What question did this study set out to answer?

To develop and evaluate a bi-timescale decision-making and control framework that balances real-time computational feasibility, safety, and efficiency for urban autonomous vehicles facing dense multi-agent interactions and occlusions.

May 29, 2026Multimodal TransportationOpen Access

Multimodal multiscale decision-making and control for urban autonomous vehicles with memory-conditioned dynamic potential field reconstruction

View Full Paper

Ask AI

Bookmark

Authors

YLYanbin Liu CZCong Zhang SCShaohua Cui

Discussion

Member takes

Overview

Experimental evaluation demonstrates reduced collision rates in urban autonomous driving, highlighting safer navigation under complex occlusions and traffic signals.

Key Points

To develop and evaluate a bi-timescale decision-making and control framework that balances real-time computational feasibility, safety, and efficiency for urban autonomous vehicles facing dense multi-agent interactions and occlusions.
Formulated the M3UDMC framework, integrating multimodal scene representation, memory-augmented risk reasoning, dynamic potential field reconstruction, and constrained model predictive control across fast and slow timescales.
Evaluated performance using high-fidelity simulations, hardware-in-the-loop testing, and real-world road experiments against Apollo 8.0, DDPG reinforcement learning, and fixed-potential MPC baselines.
M3UDMC reduced the collision rate to 7.8% across evaluated urban driving scenarios, compared to 21.3% and 18.7% for baseline methods.
Ablation studies showed that combining memory augmentation with dynamic potential field reconstruction substantially improved decision quality during occlusions, signal transitions, and rare interaction events.

Cite This Study

Liu et al. (2026) studied this question.

synapsesocial.com/papers/6a80f9b82d95cd89dee18a44 https://doi.org/10.1016/j.multra.2026.100324

View Full Paper

Ask AI

Bookmark

May 29, 2026Multimodal TransportationOpen Access

Multimodal multiscale decision-making and control for urban autonomous vehicles with memory-conditioned dynamic potential field reconstruction

View Full Paper

Ask AI

Bookmark

Authors

YLYanbin Liu CZCong Zhang SCShaohua Cui

Discussion

Member takes

Overview

Experimental evaluation demonstrates reduced collision rates in urban autonomous driving, highlighting safer navigation under complex occlusions and traffic signals.

Key Points

To develop and evaluate a bi-timescale decision-making and control framework that balances real-time computational feasibility, safety, and efficiency for urban autonomous vehicles facing dense multi-agent interactions and occlusions.
Formulated the M3UDMC framework, integrating multimodal scene representation, memory-augmented risk reasoning, dynamic potential field reconstruction, and constrained model predictive control across fast and slow timescales.
Evaluated performance using high-fidelity simulations, hardware-in-the-loop testing, and real-world road experiments against Apollo 8.0, DDPG reinforcement learning, and fixed-potential MPC baselines.
M3UDMC reduced the collision rate to 7.8% across evaluated urban driving scenarios, compared to 21.3% and 18.7% for baseline methods.
Ablation studies showed that combining memory augmentation with dynamic potential field reconstruction substantially improved decision quality during occlusions, signal transitions, and rare interaction events.

Cite This Study

Liu et al. (2026) studied this question.

synapsesocial.com/papers/6a80f9b82d95cd89dee18a44 https://doi.org/10.1016/j.multra.2026.100324

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

View Full Paper

Ask AI

Bookmark