# A-DRIVE: Autonomous Deadlock Detection and Recovery at Road Intersections for Connected and Automated Vehicles

**Authors:** Shunsuke Aoki,  Ragunathan,  Rajkumar  
**arXiv:** [2204.04910](https://arxiv.org/abs/2204.04910)  
**Date:** 2022-04-11  
**Summary page:** https://www.summarizepaper.com/en/paper/2204.04910v1/

## Key points

- The paper addresses the issue of deadlocks in connected and automated vehicles (CAVs) at road intersections, single-track lanes, and construction zones.
- The authors propose A-DRIVE, an Autonomous Deadlock Detection and Recovery Protocol.
- A-DRIVE is a decentralized and time-sensitive technique that aims to improve traffic throughput and shorten worst-case recovery time.
- A-DRIVE consists of two components: V2V communication-based A-DRIVE and Local perception-based A-DRIVE.
- V2V communication-based A-DRIVE facilitates communication between CAVs to coordinate movements and avoid blocking each other.
- Local perception-based A-DRIVE relies on local perception capabilities to detect deadlocks and recover from them.
- Simulations showed improved traffic throughput and reduced recovery time compared to a baseline protocol.
- The paper provides an analysis of different deadlock situations at stationary and dynamic intersections as well as a classification of vehicle deadlocks.
- Assumptions regarding traffic environments are discussed in the study.
- The proposed A-DRIVE protocol enhances efficiency and safety at road intersections, single track lanes, and construction zones.

## Summary

The paper titled "A-DRIVE: Autonomous Deadlock Detection and Recovery at Road Intersections for Connected and Automated Vehicles" addresses the issue of deadlocks that can occur when multiple connected and automated vehicles (CAVs) block each other at road intersections, single-track lanes, and construction zones. To tackle this problem, the authors propose A-DRIVE, an Autonomous Deadlock Detection and Recovery Protocol. A-DRIVE is a decentralized and time-sensitive technique that aims to improve traffic throughput and shorten worst-case recovery time. It consists of two components: V2V communication-based A-DRIVE and Local perception-based A-DRIVE. V2V communication-based A-DRIVE enables deadlock recovery by facilitating communication between CAVs to coordinate their movements and avoid blocking each other. On the other hand, Local perception-based A-DRIVE relies on local perception capabilities to detect deadlocks and recover from them. The authors conducted simulations to evaluate the performance of A-DRIVE compared to a baseline protocol which showed improved traffic throughput while reducing recovery time. The paper also provides an analysis of different deadlock situations at stationary and dynamic intersections as well as a classification of vehicle deadlocks. Additionally, it discusses assumptions made in the study regarding traffic environments. In conclusion, the proposed A-DRIVE protocol offers a promising solution for detecting and recovering from deadlocks in connected and automated vehicle systems by improving traffic throughput while reducing recovery time; thus enhancing the efficiency and safety of road intersections, single track lanes, and construction zones.