Mitsubishi Electric Corporation announced its development in artificial intelligence (AI) integrated automated guided vehicles (AGV). The company has developed a collaborative AI technology that uses inverse reinforcement learning (IRL) to learn and imitate the behavior of skilled workers, thereby enhancing work collaboration between humans and machines.
IRL is the main function of Mitsubishi Electric's Maisart (Mitsubishi Electric's AI created the most advanced) AI technology, which is a machine learning technology that allows machines to imitate and learn human-like actions based on a relatively small amount of data. IRL is a field in which experts or skilled workers learn their behavior by observing their compensation, goals or values. The new collaborative AI technology will be further improved through the use of prototypes, persistent field testing through test deployment, and finally in the production and distribution sites to debug AGVs and robots with humans . Mitsubishi Electric speculates that in the future, this technology will be used in autonomous vehicles and other applications.
An example of collaborative AI deployed in AGV. (Picture provided by MEC.)
Mitsubishi shares the key functions of human-machine collaboration AI
This technology is expected to improve efficiency and safety in an environment where both humans and machines work. In sites where the environment involves humans and machines, AGV will use integrated AI technology to retrieve data from the images in the video of these work areas to learn and imitate the behavior and decision-making of skilled workers. In a mixed work environment where both man and machine exist, this can help reduce dangerous and unnecessary events, such as high-risk actions, including collisions or deadlocks.
The internal simulation results conducted by Mitsubishi Electric show that compared with the traditional mixed working environment equipped with fewer intelligent machines, the operating efficiency has increased by about 30%. Since IRL requires a relatively small amount of data, the new Maisart technology reduces the amount of operational data required for learning.
Traditional machine learning is notorious for requiring a surprisingly large amount of data. In this case, if AI is to learn and imitate human actions, it needs video data. However, development using IRL reduces the amount of data required for AI to learn and imitate human behavior and decision-making. In the latest test of Mitsubishi Electric's Maisart AI, the technology only requires 10% or less of the video data required by traditional machine learning algorithms.
Cooperate with Mitsubishi Electric to develop the future of human-machine collaboration artificial intelligence
The project still requires work such as deployment and field testing. Looking ahead, Mitsubishi Electric will continue to develop its new collaborative AI for use in commercial facilities. The expected benefits include improved operational efficiency, enabling workers to distance themselves from society, and allowing machines and people to operate safely side by side in environments such as factory production lines and logistics warehouses, and autonomous vehicles. Please pay attention to the future updates and applications of IRL cooperative AI technology in AVG.
As a leading manufacturer in the global automation field, Mitsubishi Electric serves users in various industries around the world with its excellent automation products and technologies. In China, Mitsubishi Electric's PLC, inverter, servo, CNC, low-voltage electrical appliances, industrial robots, processing machines and other products operate in all walks of life. The eF@ctory intelligent manufacturing solution has built a framework for intelligent factories and implemented them on the ground. China's economic development actively contributes.