Heterogeneous computing for low-cost robotic platforms

Abstract: This study presents a novel approach to the design of cost-effective autonomous rovers using a heterogeneous computing architecture. The research focuses on the development and evaluation of an autonomous rover system, priced under $2 700 (USD), that can operate independently in various environments. The system’s performance was assessed in real-world conditions, demonstrating an average speed of 0.5 m/s, an energy consumption of 0.2 kWh, and a decision-making error rate below 5%. The results suggest that the use of heterogeneous computing architectures can lead to the development of affordable and efficient autonomous navigation systems. The paper concludes by discussing potential future research directions and the broader implications of these findings for the field of autonomous systems.

Download paper here