control - メーカー・企業と製品の一覧 | イプロス

[Data Science Use Case] Automatic Control of Heavy Machinery Using Reinforcement Learning AI

Indicating the technical possibilities for achieving business efficiency! A case study of developed control AI.

We will introduce a case study of automatic control of heavy machinery using reinforcement learning-based AI. In heavy machinery operations, there was a challenge of unstable quality among technicians due to a lack of technical inheritance accompanying the decrease in the working population. Additionally, there were significant hurdles in collecting data in heavy machinery operations, necessitating the establishment of a safe method for data collection. Therefore, we constructed a simulator that reproduces part of the heavy machinery's operations using sensor data installed on the machinery, employing machine learning techniques. We developed a control AI through reinforcement learning within that simulator environment. As a result, we were able to gain insights into the technical possibilities for improving operational efficiency. 【Case Overview】 ■Industry: Construction ■Business: Construction Work ■Challenge: Labor Reduction ■Analytics and AI Solution - Simulator Development We built a simulator that reproduces part of the heavy machinery's operations using sensor data installed on the machinery. It learns the behavior of sensor data in response to control signals for the actual machine. - Control AI Development The AI learns to minimize the error between the simulator's operation and target values in response to control signals. *For more details, please refer to the PDF document or feel free to contact us.