DeepTracker: deep learning based bio-inspired vibration source localisation device
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
This compute capsule implements DeepTracker v1, a novel bio-inspired device for vibration source localization. Integrating a spider-like sensor geometry with deep learning architectures, DeepTracker processes raw time-series data from eight onboard sensors to predict the exact location of a vibration source. Notably, the algorithm is substrate-agnostic, requiring no prior knowledge of the physical properties of the medium. Designed for real-time applicability, the inference pipeline is highly efficient, executing in just 0.005s on an NVIDIA Jetson edge platform.
Fields of Study & Subjects
Structural Health Monitoring Techniques
Machine Fault Diagnosis Techniques
Direction-of-Arrival Estimation Techniques
Publication details
- Journal / source
- Code Ocean
- Publication date
- 1 January 2026
- Language
- EN