Hitachi combines control engineering with AI and software engineering “Physical AI” to develop technology to improve development efficiency and reuse of control software

Hitachi has developed technology aimed at realizing “Physical AI,” which combines control engineering with AI (Artificial Intelligence) and software engineering for control software in the automotive and logistics fields. In the automotive field, by incorporating API (Application Programming Interface) information specific to the actual controller into generative AI , the automatic generation of test scripts for the actual device, which was previously difficult, was achieved, reducing integration testing man-hours by 43%. In the logistics field, by analyzing variables in the on-site environment and work and reflecting them in the architecture design, the reusability of autonomous robot control software and on-site work efficiency were improved. These technologies will contribute to the realization of sustainable social infrastructure by improving the efficiency of control software development and reducing the burden on on-site workers.

■ Features of the technologies and solutions developed to solve the issues

To address a wide range of on-site challenges, Hitachi has developed technology aimed at realizing “Physical AI,” which combines control engineering with AI and software engineering.

Also Read: Metareal AM Launches Premium Plan for AI Agri Equipment Analysis

Hitachi and Astemo Inc. have developed test generation AI technology for in-vehicle systems. By incorporating API information specific to the actual controller and on-site know-how into the generative AI (large-scale language model + search expansion generation), integrated test scripts for the actual device that reflect on-site knowledge are automatically generated from test case specifications written in natural language. This has streamlined the creation of test scripts, which previously required a significant amount of man-hours, and achieved a 43% reduction in man-hours in a pilot project for multi-core ECU (Electronic Control Unit) integration testing. It can also flexibly accommodate on-site hardware configurations, enabling highly reliable AI utilization.

Case 2: Technology to improve reusability of autonomous robot control software through variability management in the logistics field 

Hitachi has developed a variability management technology that enables flexible management through software by analyzing in advance the various variables that occur in products, environments, and work content at factories, logistics centers, and other work sites, and organizing them into functional models. By modularizing the robot control software and implementing it as a node that runs on ROS (Robot Operating System)*11, it is possible to quickly respond to new products and changes in picking/placing conditions, improving software reusability. Through interviews with on-site engineers and robot operators and demonstration experiments, we confirmed that the technology improves the efficiency of system configuration work.

SOURCE: PRTimes