Kintetsu Railway Co., Ltd. (HQ: Tennoji-ku, Osaka; President: Yasushi Hara; hereafter 'Kintetsu') and Sharp Corporation (HQ: Chuo-ku, Osaka; President: Tetsuji Kawamura; hereafter 'Sharp') will jointly begin a pilot test for a railway underfloor inspection system utilizing image processing technology on May 26, 2026.

The two companies began joint research in the fall of 2020 with the aim of maintaining and improving inspection accuracy while reducing workload, a critical aspect of safe railway operation. They have been developing this system by combining Kintetsu's maintenance know-how with Sharp's image analysis and processing technology, developed through its TV and smartphone business.

This system analyzes images captured by cameras to automatically check for abnormalities in underfloor equipment.

Building on findings from this pilot test, the companies will strive for early commercialization and aim to utilize the system in the future for increased inspection efficiency and predictive maintenance through human-system collaboration.

1. Overview In this pilot test, wheel sensors, cameras, and lighting equipment will be installed at the entrance to the inspection track at Saidaiji Depot (Nara City, Nara Prefecture). Upon detecting the train's entry, the system automatically photographs the underfloor equipment. By linking images with data from ID tags on the trains, the system analyzes and lists abnormalities in equipment and wear conditions of brake shoes. Verification of these operations and accuracy will confirm the system's effectiveness in streamlining inspection work.

2. Duration From Tuesday, May 26, 2026, for the time being.

3. Verification Items (1) Accuracy verification of inspection functions. (2) Accuracy verification under real-world conditions involving changes in weather, sunlight, and vehicle status. (3) Verification of system operational methods.

4. Inspection Examples - Brake shoe remaining thickness measurement: Measures brake shoe thickness to determine if replacement is necessary. - Equipment detection: Detects whether equipment is in the correct position. - Equipment deformation detection: Detects deformation of equipment or the orientation of equipment boxes and cock handles.

FACT BOX

  • Source: PR TIMES
  • Category: Event