Abstract：A quantum-inspired optimization system, the Simulated Bifurcation Machine (SBM), has been integrated into a mobile robot, enabling real-time multi-object tracking on compact, low-power devices. This breakthrough promises faster, more efficient autonomous navigation for robots and vehicles across logistics and smart mobility applications.
A significant advance in autonomous mobility has been achieved with the integration of the Simulated Bifurcation Machine (SBM) into a mobile robot. For the first time, a quantum-inspired optimization system has been embedded directly into a mobile platform, allowing real-time processing of complex tasks like multi-object tracking while meeting strict size and power constraints. This innovation enhances robot vision systems, which are essential for autonomous vehicles and smart mobility applications.
SBM operates on conventional hardware such as FPGAs, GPUs, and ASICs, using quantum-inspired algorithms to solve complex combinatorial problems. In prototype testing, the system demonstrated the ability to track multiple objects in real time, even in crowded environments with occlusions, achieving a 4% increase in tracking accuracy and a 23% improvement in handling partially obscured objects. Its high-speed search capability allows the robot to maintain continuous tracking of both one-to-one and one-to-many object relationships.
The system also achieved 23 frames per second, surpassing the typical 10 FPS standard for automated driving, while remaining compact and energy-efficient. By performing real-time optimization on embedded devices, SBM overcomes traditional limitations that required large, high-performance servers for such tasks. In real-world trials, the robot successfully executed dynamic path planning, avoiding obstacles while improving navigation efficiency using SBM’s real-time tracking data.
Figure: Embedded platform and autonomous mobile robot equipped with SBM
This breakthrough demonstrates the practical application of quantum-inspired computing in autonomous mobility and sets the stage for future innovations. Potential uses include automated vehicles, multi-robot coordination, complex route optimization, and real-time task allocation, offering transformative benefits for industries that rely on autonomous systems.
