Zero-Knowledge Proof-Based Verification System Based on Environmental Sensing for Reliable Operation of AI-Driven Autonomous Robots

Abstract

The smart cities that collaborate with AI-driven autonomous robots are attracting attention for supporting various social activities in the real world. In facilities that provide such services, various systems managing the facility and robots may coexist in the common area. By enabling the systems to interoperate and share information about the status of the facility and robots, it becomes possible to realize a variety of services that support safety and security within the facility. However, while the operators of the facility want to monitor the detailed conditions of the robots, the operators of the robots are cautious about providing the information about the status of robots such as the moving trajectory and various sensor data. To resolve this dilemma, we propose a new system that enables estimation of the operational conditions of the robots by verifying the positions and trajectories at landmarks in the facility without disclosing their internal information. In the proposed system, we focus on the observation of environmental information that accurately reflects the real-world situation for estimating the proximity between the robot and each landmark. As the environmental information, both systems on robots and a facility measure CSI (Channel State Information) and acoustic information. In addition, by utilizing zero-knowledge proof (ZKP) technology, the system for the facility confirms the reliability of the process for estimating the proximity of the robots to the landmark without exchanging detailed internal information. Through the proof-of-concept experiment, applying the proposed system achieved high-accuracy proximity detection with both methods (CSI and acoustic information) yielding precision and recall rates exceeding 0.90.