Japan's elevator industry is witnessing a surge in maintenance demand while the skilled technician workforce is shrinking as many reach retirement age.
To address this challenge, Hitachi Building Systems—a subsidiary of the Hitachi Group—has selected the open-source multimodal AI platform from technology giant Nvidia and customized the model by training it with the product's technical documents and maintenance manuals. This AI system is capable of providing voice guidance and generating video-based reports.

A Hitachi Building Systems technician inspecting an elevator.
Specifically, during elevator inspections, technicians will be equipped with an edge server near the work area to perform data processing. This server connects to the data center via a wireless network. Technicians use cameras integrated into safety helmets or smart glasses, equipped with microphones and speakers, allowing the AI to interact in real-time.
During the inspection process, the AI provides step-by-step voice guidance, for example: “Use the key to open the elevator door” or “Check the cabin position.” The system is also capable of detecting incorrect operations and issuing warnings.
If the technician detects a component showing signs of needing replacement, they can take a photo for the AI to analyze; the system then evaluates the image and provides the corresponding component part number for replacement.
The entire inspection process is recorded and sent to the data center. Based on data from this video, the system will automatically generate a work report—the AI converts video content into text and attaches necessary images to the report.
This AI technology is expected to be deployed in the second half of Fiscal Year 2026, marking a significant step in Hitachi's digital transformation strategy for building maintenance operations. (The fiscal year in Japan begins on April 1 and ends on March 31 of the following year).
Mr. Noriharu Amiya, President of Hitachi Building Systems, stated that Hitachi's goal extends beyond elevators, aiming to expand the application of this system to the maintenance of the entire building.
According to data from the Japan Elevator Association (JEA), in Fiscal Year 2024, approximately 782,200 elevators underwent periodic maintenance nationwide, an increase of over 30% compared to two decades ago. This reflects a strong demand for automation and AI solutions in the elevator maintenance sector amidst a shortage of skilled labor.
https://tapchithangmay.vn/hyundai-phat-trien-thang-may-mo-dun-giam-tai-nan-va-rut-ngan-thoi-gian-lap-dat/




























