Summary of Savvy: Trustworthy Autonomous Vehicles Architecture, by Ali Shoker et al.
Savvy: Trustworthy Autonomous Vehicles Architecture
by Ali Shoker, Rehana Yasmin, Paulo Esteves-Verissimo
First submitted to arxiv on: 8 Feb 2024
Categories
- Main: Artificial Intelligence (cs.AI)
- Secondary: Systems and Control (eess.SY)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The increasing interest in Autonomous Vehicles (AVs) due to business, safety, and performance reasons has led to recent success in AV architectures relying on AI models. However, despite this progress, fatal incidents have hindered the widespread adoption of full AVs. To address this issue, a new trustworthy intelligent AV architecture called Savvy is proposed. Savvy achieves the best of both worlds by separating the control plane and data plane, ensuring safety-first principles are applied. The control plane assumes control using design-time defined rules, while the data plane optimizes decisions within safety time-bounds through guided Time-aware predictive quality degradation (TPQD). For example, Savvy can safely identify an elephant as an obstacle earlier rather than recognizing it too late. This paper presents the motivations and concept of Savvy, with empirical evaluation ongoing. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Autonomous Vehicles are getting more popular because they’re safer and better for business. But sometimes these cars have accidents that hurt people. To make them safer, we need to rethink how we build them. One way to do this is by using artificial intelligence (AI) in a special way. We propose an architecture called Savvy that separates the parts of the car that control it from the parts that make decisions. This helps ensure safety comes first. The controller uses rules set ahead of time, while the decision-maker optimizes choices within safe time limits. For example, Savvy could recognize an elephant as a hazard earlier rather than later. This idea is still being developed and tested. |
