Toshiba Corporation and MIRISE Technologies Corporation have achieved the world’s first deployment of a quantum-inspired optimization computer inside an autonomous mobile robot.
The milestone, demonstrated in Japan, marks a significant leap in real-time decision-making for self-driving systems operating under strict size, cost, and power constraints.
By embedding Toshiba’s proprietary Simulated Bifurcation Machine (SBM) directly into a mobile platform, the companies have proven that advanced optimization algorithms can function efficiently on compact, low-power hardware.
What Is Quantum-Inspired Optimization?
Quantum-inspired optimization computers use mathematical algorithms derived from quantum computing principles to solve complex combinatorial optimization problems at high speed.
READ ALSO: https://modernmechanics24.com/post/boomerang-quakes-seismic-threat-explain/
Unlike true quantum computers, these systems do not require specialized quantum hardware. They operate on conventional platforms such as FPGAs, GPUs, and ASICs.
This makes them highly attractive for industrial applications, especially in robotics and autonomous vehicles, where real-time processing and efficiency are critical.
Industries face growing labor shortages. As a result, demand for self-driving vehicles and autonomous mobile robots is accelerating across logistics, smart mobility, manufacturing, and infrastructure management.
These systems must continuously detect obstacles, track objects, and plan routes in dynamic environments, all within tight control cycles.
However, balancing complex computational demands with strict limitations on power consumption, device size, and cost has posed a serious technical hurdle.
Toshiba addressed this challenge by developing a new multi-object tracking algorithm powered by its SBM quantum-inspired optimization engine.
The company then implemented the solution on an embedded FPGA using proprietary circuit design technology. MIRISE integrated the FPGA into its autonomous mobile robot platform and successfully demonstrated real-time navigation in live conditions.
WATCH ALSO: https://modernmechanics24.com/post/byd-seal-smooth-electric-power/
This marks the first time a quantum-inspired optimization computer has been embedded directly into a mobile robotic system for autonomous control.
At the core of the innovation is a new algorithm capable of tracking multiple moving objects such as people, vehicles, and obstacles, even when they cross paths or temporarily obscure each other.
Traditional tracking systems often struggle when objects overlap or briefly go out of view. It leads to misidentification or tracking failures. Toshiba’s solution formulates multi-object tracking as a combinatorial optimization problem. While conventional approaches rely on one-to-one matching between detected and tracked objects, the SBM-powered algorithm enables large-scale, high-speed searches that identify potential one-to-many matchings.
This capability allows the system to re-track objects after occlusion and improve motion prediction accuracy.
Performance evaluations using the Higher Order Tracking Accuracy (HOTA) metric showed a 4 percent improvement over standard tracking benchmarks and a 23 percent improvement in scenarios specifically designed to test object obscuration.
READ ALSO: https://modernmechanics24.com/post/womb-transplants-from-deceased-donor/
These gains translate into safer, smoother navigation in crowded, complex environments. It is a critical requirement for the real-world deployment of automated systems.
The SBM architecture is designed for high parallelism, enabling multiple computations to run simultaneously. Leveraging this characteristic, Toshiba achieved high-speed processing even within the constraints of embedded FPGA hardware.
The system demonstrated multi-object detection and tracking at 23 frames per second (FPS), more than double the 10 FPS typically required for automated driving applications. This performance enables advanced optimization tasks — previously restricted to high-performance servers — to operate in real time on compact robot control units and in-vehicle systems.
To validate the technology, Toshiba and MIRISE mounted the FPGA equipped with the SBM algorithm onto an autonomous mobile robot developed by MIRISE. During testing, the robot successfully performed dynamic path planning while avoiding multiple moving obstacles.
MIRISE enhanced the robot’s navigation capabilities by incorporating SBM tracking data into its path-planning mechanism. By analyzing positional confidence and movement direction, the system dynamically adjusted object occupancy areas and predicted future positions. This reduced unnecessary avoidance maneuvers and ensured smoother, more efficient movement.
WATCH ALSO: https://modernmechanics24.com/post/china-40000t-sichuan-assault-ship/
The robot uses a combination of cameras, sensors, and the SBM-based tracking algorithm. This helps select optimal routes in real time, even in environments with both static and moving obstacles.
The companies plan to expand the application of embedded quantum-inspired optimization to a wider range of autonomous control systems. Future developments include cooperative control of multiple mobile robots, advanced route optimization in complex environments, and real-time task allocation.
Apart from robotics, the technology could support factory transport robots, warehouse automation, autonomous construction and agricultural machinery, smart city infrastructure, and energy management systems.
By demonstrating that quantum-inspired computing can operate efficiently inside compact mobile platforms, Toshiba and MIRISE have opened a new chapter in autonomous system design. It will showcase how powerful optimization capabilities move from centralized servers directly into the machines navigating the real world.
