Integrating Modular Pipelines with End-to-End Learning: A Hybrid Approach for Robust and Reliable Autonomous Driving Systems

Luis Alberto Rosero,Iago Pachêco Gomes,Júnior Anderson Rodrigues da Silva,Carlos André Przewodowski,Denis Fernando Wolf,Fernando Santos Osório

DOI: https://doi.org/10.3390/s24072097

IF: 3.9

2024-03-26

Sensors

Abstract:Autonomous driving navigation relies on diverse approaches, each with advantages and limitations depending on various factors. For HD maps, modular systems excel, while end-to-end methods dominate mapless scenarios. However, few leverage the strengths of both. This paper innovates by proposing a hybrid architecture that seamlessly integrates modular perception and control modules with data-driven path planning. This innovative design leverages the strengths of both approaches, enabling a clear understanding and debugging of individual components while simultaneously harnessing the learning power of end-to-end approaches. Our proposed architecture achieved first and second place in the 2023 CARLA Autonomous Driving Challenge's SENSORS and MAP tracks, respectively. These results demonstrate the architecture's effectiveness in both map-based and mapless navigation. We achieved a driving score of 41.56 and the highest route completion of 86.03 in the MAP track of the CARLA Challenge leaderboard 1, and driving scores of 35.36 and 1.23 in the CARLA Challenge SENSOR track with route completions of 85.01 and 9.55, for, respectively, leaderboard 1 and 2. The results of leaderboard 2 raised the hybrid architecture to the first position, winning the edition of the 2023 CARLA Autonomous Driving Competition.

engineering, electrical & electronic,chemistry, analytical,instruments & instrumentation

What problem does this paper attempt to address?

### What problem does this paper attempt to solve?

This paper aims to address the limitations of both modular and end-to-end approaches in autonomous driving systems and proposes a hybrid architecture that combines the advantages of both.

1. **Problems with the modular approach**:
- While the modular approach offers interpretability and ease of debugging, its complex component coupling leads to increased error propagation, making the overall architecture complex and costly to maintain.

2. **Problems with the end-to-end approach**:
- The end-to-end approach simplifies the system architecture and reduces the need for manual feature engineering, but its decision-making process lacks transparency, requires a large amount of training data, and has limited generalization ability in different driving scenarios.

### Proposed solution in the paper:

The paper proposes a hybrid architecture that combines modular perception and control components with data-driven path planning. This hybrid architecture aims to:
- Combine the interpretability of modular systems with the learning capability of end-to-end methods.
- Improve the robustness and reliability of the system.
- Perform well in both map-based and mapless navigation scenarios.

### Experimental validation:

The hybrid architecture achieved significant results in the 2023 CARLA Autonomous Driving Challenge, securing first place in the MAP track and second place in the SENSOR track. The specific results are as follows:
- In the MAP track, it achieved a driving score of 41.56 and a highest route completion rate of 86.03%.
- In the SENSOR track, it achieved a driving score of 35.36 and a route completion rate of 85.01%.

These experimental results demonstrate the effectiveness of the hybrid architecture in different driving scenarios.