Patrol inspections of equipment and instruments are essential for facility maintenance. However, visual inspections undeniably place a significant burden on inspectors. The key challenge is to reduce this burden while digitizing atypical analog information within the equipment to improve efficiency.

Japan's LiLz Gauge and LiLz Guard, developed by LiLz, address this challenge. Designed specifically for maintenance, these products streamline daily patrol inspections. Let's take a closer look at their key features and benefits.

A Cloud Service for Remote Visual Inspections in Equipment Maintenance

We are the Thai subsidiary of Shinko Corporation, a member of the Kobe Steel Group. In addition to handling steel products, non-ferrous products, welding materials, and machinery, we are also focused on improving on-site productivity and efficiency, ensuring safety through maintenance, and building factory automation (FA) systems. Through our daily operations, we have come to recognize the various challenges faced in the maintenance activities of the manufacturing industry.

Even in Japan, many sites still rely on visual inspections. According to a survey conducted by LiLz, based on the Common Building Maintenance Specifications issued by Japan's MLIT (Ministry of Land, Infrastructure, Transport and Tourism), 87% of inspections are still dependent on visual checks. Many locations face challenges in securing power and network connectivity, which has slowed the adoption of IoT. With limited personnel available for inspections, routine checks are often delayed. However, inspections conducted less than once a day provide insufficient data for informed decision-making. Patrols and maintenance activities during severe weather, such as typhoons, are demanding, and the transfer of maintenance skills to the next generation is also facing challenges.

Contributing to the resolution of these challenges is LiLz Gauge , a product from LiLz that we began selling in June 2020. This cloud service uses low-power IoT cameras and machine learning to easily enable the remote monitoring of visual patrol inspections for analog meters and other instruments. The system operates without the need for a power supply, wiring, or network installation, with a battery life of 3 years while capturing images three times a day. Since it runs solely on LTE, there’s no need to secure an additional gateway, and it can also be installed outdoors. LiLz Gauge supports a wide range of instruments, automatically reading multiple meters from a single image. By using APIs, users can obtain data such as "instrument readings" and "images captured by the camera," allowing integration with existing equipment management systems.

Datafication of Atypical Information for Inspection Monitoring and Abnormality Detection

In addition, starting in late October 2024, we launched “LiLz Guard”, a new product that combines IoT and AI technologies. This cloud service includes a feature that quantifies the "degree of deviation from normal" in captured site images using AI. By simply registering an image of the normal state, the system can detect abnormalities based on differences observed in subsequent images.

▲ Sample of LiLz Guard setting screen

There are many sites where inspection and monitoring are essential, such as chemical leaks from tanks, water leaks from air conditioning systems, belt conveyor damage, and monitoring of equipment surroundings during disasters. However, it is difficult to quantify such atypical information. By capturing images with LiLz's IoT cameras and analyzing them with AI, measurement data can be integrated with sensor information. When certain thresholds are exceeded or deviations from past data trends are detected, the system triggers alerts. Additionally, machine learning enables the prediction of potential equipment failures.

▲ Sample of the LiLz Guard management screen: Abnormalities are notified with images for easy understanding

For example, in the visual inspection of a lubricant oil tank, standard information such as oil level, internal tank pressure, and oil temperatures at the inlet and outlet can be quantified and digitized. By installing LiLz Guard on roof drains, it becomes possible to periodically monitor clogs caused by fallen leaves and prevent drainage blockages before they occur.

▲ Visualization of abnormalities such as “leakage,” “clogging,” “intrusion,” and “collapse” within facilities

While there are AI services that detect abnormal behavior of workers for security purposes or identify defective products on production lines, LiLz Guard is an AI service specifically developed for visual inspections in maintenance. Its ability to detect maintenance-related abnormalities that were previously undetectable is a significant advantage, setting it apart from other solutions.

It is also highly user-friendly, offering simple and intuitive operation. The language is designed to be easily understood by on-site workers. Since the results are consistent regardless of who uses it, it is especially suited for sites dealing with maintenance staff shortages and difficulties in retaining personnel.

LiLz Guard can also be retrofitted to sites that already have cameras installed. Please utilize this solution to address maintenance challenges that were previously unsolvable.