AI Research in Japanese Enterprises – Toshiba and Mitsubishi


As the other usual suspects in Japan are moving towards the artificial intelligence market, Toshiba is not sitting idle. Like other large companies in Japan, it has many subsidiaries and related companies. Toshiba Solutions (hereinafter, Toshiba-Sol) is a wholly owned subsidiary of Toshiba Corporation.

Toshiba Solutions Logo

Dialogue System for Inheritance Advice

Several companies in the US and Japan have produced dialogue-based virtual assistants, like IBM Watson. Toshiba-Sol worked with the Tohou Bank to produce such a system. It claims that its system interacts with customers directly via text and voice. The dialogue is specialized for inheritance advice.

On the other hand, the application of IBM Watson (at call centers of Sumitomo Mitsui Banking Corp. and Mizuho Bank) will be indirect. In other words, each customer will talk to an operator who will then interact with Watson to get an answer. Finally, the operator will give the answer to the customer.

Humanoid Robot by Toshiba

Toshiba displayed a humanoid robot in April that provides information about the Mitsukoshi department store.

Aiko, the robot by Toshiba


Mitsubishi Electric has produced two AI technology applications.

Driver Support to Avoid Accidents

Mitsubishi Electric developed a system to detect unreliable driving by observing steering handling, obtaining face position and heart rate, and warning the driver.

Group Control System

A large building has many elevators. Do you ever wonder how those elevators are managed? What does it take to move an elevator to the right floor most efficiently? Their algorithm is unique in minimizing people’s frustration in waiting for the elevator to come to the floor they are on. They found out that frustration grows in proportion to the square of waiting time.

Mitsubishi’s algorithm uses its AI technology to optimize elevator management by predicting the traffic pattern of a given building. Moreover, the next step of the group control system is a “destination prediction system.” This means that when you are on the first floor and want to go to tenth floor, you will be notified which elevator to take when you push a button for the destination floor. In a typical elevator, you do not select the destination floor until you enter the elevator. This may not sound like a big deal, but it can keep the movements of elevators to a minimum, leading to the saving of electricity and human traffic within a building.


My impression of the AI market is that each company has its own solution, but they all seem to be a single tool or a single tool set. I wonder if this approach works better than a platform-based approach like IBM Watson’s. It remains to be seen. I will keep reporting each company’s AI efforts as new information becomes available.

Zen Kishimoto

About Zen Kishimoto

Seasoned research and technology executive with various functional expertise, including roles in analyst, writer, CTO, VP Engineering, general management, sales, and marketing in diverse high-tech and cleantech industry segments, including software, mobile embedded systems, Web technologies, and networking. Current focus and expertise are in the area of the IT application to energy, such as smart grid, green IT, building/data center energy efficiency, and cloud computing.

, , ,

No comments yet.

Leave a Reply