Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Dodging the Moose: Experimental Insights in Real-Life Automated Collision Avoidance

Leila Gharavi, Simone Baldi, Yuki Hosomi, Tona Sato, Bart De Schutter, Binh-Minh Nguyen, Hiroshi Fujimoto

TL;DR

The paper tackles real-time collision avoidance after sudden static obstacle appearance in automated driving, focusing on the moose-test emergency scenario. It demonstrates that state-of-the-art nonlinear MPC often fails to find feasible solutions within tight computation budgets and proposes a human-like Maximum Steering Feed-forward (MSF) planner to assist when MPC is infeasible or poor. By combining MPC with MSF and using a sigmoid barrier for safety in an integrated planning framework, the authors validate the approach on a real electric vehicle (FPEV2-Kanon), showing safer, smoother evasive maneuvers across multiple speeds and obstacle configurations while outperforming a standalone MPC planner. The work advances practical real-time collision avoidance by leveraging physics-based planning and human-inspired reflexes, with potential impact on safety-critical autonomous driving applications.

Abstract

The sudden appearance of a static obstacle on the road, i.e. the moose test, is a well-known emergency scenario in collision avoidance for automated driving. Model Predictive Control (MPC) has long been employed for planning and control of automated vehicles in the state of the art. However, real-time implementation of automated collision avoidance in emergency scenarios such as the moose test remains unaddressed due to the high computational demand of MPC for evasive action in such hazardous scenarios. This paper offers new insights into real-time collision avoidance via the experimental imple- mentation of MPC for motion planning after a sudden and unexpected appearance of a static obstacle. As the state-of-the-art nonlinear MPC shows limited capability to provide an acceptable solution in real-time, we propose a human-like feed-forward planner to assist when the MPC optimization problem is either infeasible or unable to find a suitable solution due to the poor quality of its initial guess. We introduce the concept of maximum steering maneuver to design the feed-forward planner and mimic a human-like reaction after detecting the static obstacle on the road. Real-life experiments are conducted across various speeds and level of emergency using FPEV2-Kanon electric vehicle. Moreover, we demonstrate the effectiveness of our planning strategy via comparison with the state-of- the-art MPC motion planner.

Dodging the Moose: Experimental Insights in Real-Life Automated Collision Avoidance

TL;DR

The paper tackles real-time collision avoidance after sudden static obstacle appearance in automated driving, focusing on the moose-test emergency scenario. It demonstrates that state-of-the-art nonlinear MPC often fails to find feasible solutions within tight computation budgets and proposes a human-like Maximum Steering Feed-forward (MSF) planner to assist when MPC is infeasible or poor. By combining MPC with MSF and using a sigmoid barrier for safety in an integrated planning framework, the authors validate the approach on a real electric vehicle (FPEV2-Kanon), showing safer, smoother evasive maneuvers across multiple speeds and obstacle configurations while outperforming a standalone MPC planner. The work advances practical real-time collision avoidance by leveraging physics-based planning and human-inspired reflexes, with potential impact on safety-critical autonomous driving applications.

Abstract

The sudden appearance of a static obstacle on the road, i.e. the moose test, is a well-known emergency scenario in collision avoidance for automated driving. Model Predictive Control (MPC) has long been employed for planning and control of automated vehicles in the state of the art. However, real-time implementation of automated collision avoidance in emergency scenarios such as the moose test remains unaddressed due to the high computational demand of MPC for evasive action in such hazardous scenarios. This paper offers new insights into real-time collision avoidance via the experimental imple- mentation of MPC for motion planning after a sudden and unexpected appearance of a static obstacle. As the state-of-the-art nonlinear MPC shows limited capability to provide an acceptable solution in real-time, we propose a human-like feed-forward planner to assist when the MPC optimization problem is either infeasible or unable to find a suitable solution due to the poor quality of its initial guess. We introduce the concept of maximum steering maneuver to design the feed-forward planner and mimic a human-like reaction after detecting the static obstacle on the road. Real-life experiments are conducted across various speeds and level of emergency using FPEV2-Kanon electric vehicle. Moreover, we demonstrate the effectiveness of our planning strategy via comparison with the state-of- the-art MPC motion planner.
Paper Structure (19 sections, 20 equations, 10 figures, 2 tables)

This paper contains 19 sections, 20 equations, 10 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Vehicle Model
  3. Kinematic Model
  4. Equations of Motion
  5. Proposed Control System
  6. Perception
  7. Planner
  8. Controller
  9. Design of the Planning Strategy
  10. MSF Planner
  11. MPC Planner
  12. FBL Controller
  13. System Implementation
  14. Results and Discussion
  15. Relaxed cases with $v_x = 5$m/s
  16. ...and 4 more sections

Figures (10)

  • Figure 1: Elements of the collision avoidance problem after detecting a static obstacle
  • Figure 2: Single-track vehicle model
  • Figure 3: The proposed control architecture
  • Figure 4: Maximum steering maneuver (\ref{['eq:ymax']}) for different longitudinal velocities.
  • Figure 5: Sigmoid barrier function examples
  • ...and 5 more figures

Theorems & Definitions (5)

  • Remark 1
  • Example 1
  • Example 2
  • Remark 2
  • Remark 3