Organizations: Department of Electrical and Computer Engineering, Carnegie Mellon Universty · Faculty of Medicine, King Mongkut’s Institute of Technology Ladkrabang, Thailand · Department of Mechanical Systems Engineering, Nagoya University, Japan
Ensuring safe autonomous driving in the presence of occlusions poses a significant challenge in its policy design. While existing model-driven control techniques based on set invariance can handle visible risks, occlusions create latent risks in which safety-critical states are not observable. Data-driven techniques also struggle to handle latent risks because direct mappings from risk-critical objects in sensor inputs to safe actions cannot be learned without visible risk-critical objects. Motivated by these challenges, in this paper, we propose a probabilistic safety certificate for latent risk. Our key technical enabler is the application of probabilistic invariance: It relaxes the strict observability requirements imposed by set-invariance methods that demand the knowledge of risk-critical states. The proposed techniques provide linear action constraints that confine the latent risk probability within tolerance. Such constraints can be integrated into model predictive controllers or embedded in data-driven policies to mitigate latent risks. The proposed method is tested using the CARLA simulator and compared with a few existing techniques. The theoretical and empirical analysis jointly demonstrate that the proposed methods assure long-term safety in real-time control in occluded environments without being overly conservative and with transparency to exposed risks.
Figures & tables
Fig. 1 : Visually occluded intersection scenario of interest.
Fig. 2 : CARLA Scenario Setup.
Fig. 3 : Estimation of safety probability Ψ(x) with grid discretization dxinit=2 for initial position and dvinit=0.5 for initial velocity. Blue curves show the typical vehicle trajectories under the proposed safe controller, with the blue dots as initial states.
Fig. 4 : Vehicle velocities of the proposed safe controller and baselines methods. Triangle indicates safely pass through the intersection. Square indicates collision with pedestrian.
Fig. 5 : Safety-efficiency trade-offs.
Control Method
Settings
1−ϵ
Psafe↑
t(s)↓
Proposed
0.9
0.98
26.94
Worst-Case
-
1
55.49
Planning-based
xinit=−180m
-
0.96
34.64
OA-MPC
vinit=2m/s
-
0.94
38.29
TransFuser
-
0.53
33.76
PID
-
1
32.68
TABLE I : Empirical safety probabilities and traveling time. ↑ and ↓ indicate larger or smaller values preferred, respectively.
Fig. 6 : Visualization of the scenario and the corresponding processed sensor inputs for TransFuser [ 61 ] . (a) Birds’ eye view of the scenario. (b) HD map prediction. (c) Obstacle prediction. (d) LiDAR ground panel. (e) First-person view from the ego vehicle. (f) Predicted depth. (g) Predicted semantics.
Fig. 7 : Vehicle velocities of the proposed safe controller with different α . In all cases the vehicle safely pass through the intersection (indicated by the triangle).
α(h)
Psafe↑
t(s)↓
0.05h
0.96
19.80
0.10h
1.00
20.95
0.20h
1.00
20.25
0.50h
1.00
22.05
1.00h
0.98
23.30
TABLE II : Empirical safety probability and average traveling time with different α . ↑ and ↓ indicate larger or smaller values preferred, respectively.
Control Method
Settings
1−ϵ
Psafe↑
t(s)↓
Proposed
0.95
1
25.36
Worst-Case
-
1
46.98
Planning-based
xinit=−180m
-
0.94
32.98
OA-MPC
vinit=5m/s
-
0.86
41.60
TransFuser
-
0.58
31.8
PID
-
0.7
13.16
TABLE III : Empirical safety probabilities and traveling time. ↑ and ↓ indicate larger or smaller values preferred, respectively.
Fig. 8 : Ego vehicle velocity under different pedestrian occurrence probabilities. In all cases the vehicle safely pass through the intersection (indicated by the triangle).
Fig. 9 : Safe control actions under different pedestrian occurrence probabilities.
Autonomous driving develops vehicles that perceive road scenes, predict interactions, and plan and control safe movement with limited or no human input.
Occlusion-aware prediction remains a critical challenge in autonomous driving due to the inherent uncertainty of unobserved regions. Existing approaches either overestimate risk based on reachable states or struggle to predict accurate trajectories under high occlusion uncertainty. To address these limitations, we propose a unified risk map modeling and learning framework for partially observable environments. Our method integrates traffic flow risk and collision risk through spatiotemporal modeling, enabling fine-grained assessment of occlusion-induced hazards. To address the scarcity of scenarios involving occluded interactions, we introduce a diffusion-based scenario generation framework that produces realistic yet adversarial scenarios. We integrate the modeling and learning of a unified risk map into a framework that supports risk-aware planning under partial observability. Experiments on the Waymo Open Motion Dataset show that our method significantly outperforms the state-of-the-art occlusion-aware baseline, improving minimum time-to-collision by 0.78 times and average time-to-collision by 1.67 times. The proposed framework offers a comprehensive and practical solution for risk-aware planning in partially observable environments.
Jie Jia, Yaofeng Su, Zeyu Bao +4
Fudan University, China · Tongji University, China
Ensuring safe driving while maintaining travel efficiency for autonomous vehicles in dynamic and occluded environments is a critical challenge. This paper proposes an occlusion-aware contingency safety-critical planning approach for real-time autonomous driving. Leveraging reachability analysis for risk assessment, forward reachable sets of phantom vehicles are used to derive risk-aware dynamic velocity boundaries. These velocity boundaries are incorporated into a biconvex nonlinear programming (NLP) formulation that formally enforces safety using spatiotemporal barrier constraints, while simultaneously optimizing exploration and fallback trajectories within a receding horizon planning framework. To enable real-time computation and coordination between trajectories, we employ the consensus alternating direction method of multipliers (ADMM) to decompose the biconvex NLP problem into low-dimensional convex subproblems. The effectiveness of the proposed approach is validated through simulations and real-world experiments in occluded intersections. Experimental results demonstrate enhanced safety and improved travel efficiency, enabling real-time safe trajectory generation in dynamic occluded intersections under varying obstacle conditions. The project page is available at https://zack4417.github.io/oacp-website/.
Lei Zheng, Rui Yang, Minzhe Zheng +3
Robotics and Autonomous Systems Thrust, The Hong Kong University of Science and Technology (Guangzhou), Guangzhou 511453, China · School of Engineering, Great Bay University, Dongguan 523000, China · Cheng Kar-Shun Robotics Institute, The Hong Kong University of Science and Technology, Hong Kong SAR, China
The deployment of autonomous vehicles in urban environments introduces significant safety challenges, particularly in scenarios with occlusions, where critical traffic participants may be hidden from view. Recent accidents involving driverless vehicles highlight the importance of motion planners that explicitly addresses the risks posed by occlusions. In this work, we propose a formal, occlusion-aware trajectory planning framework that guarantees collision avoidance even when there are possible hidden traffic participants. Building on our previous methods that apply reachability analysis to sequentially determine the possible states of hidden traffic participants, we integrate a tree-based motion planner capable of reasoning over future observations and the absence thereof. This approach reduces conservativeness while maintaining safety guarantees. We demonstrate the effectiveness of our framework in a challenging simulated occluded scenario, showing that it pro-actively and efficiently guarantees collision-avoidance.
Truls Nyberg, Anna Gautier, Jana Tumova
KTH Royal Institute of Technology · Traton AB · Chalmers University of Technology +1