This Japanese IoT Company Uses Computer Vision And Machine Learning To Automate Daily Visual Inspections In Locations With No Power Nor Network Supply

Facility management organizations must perform visual inspections daily, which must devote significant time to visiting all inspected sites. These companies want to increase their productivity by adopting IoT devices, but existing IoT solutions don’t satisfy their needs because most field locations lack power and network access.

LiLz Gauge is a cloud service that enables remote visual inspections of analog gauges using ultra-low power consumption IoT cameras and machine learning. Many machine learning-based systems necessitate extensive data preprocessing beforehand. LiLz Gauge is a deep learning-based computer vision system that doesn’t require consumers to annotate enormous volumes of data.

In regions with no electricity or network, LiLz Gauge automates daily visual examinations. LiLz Gauge greatly reduces the time necessary for visual inspections with its LiLz Cam. LiLz Cam are cameras with built-in LTE connectivity, three years of battery life, and AI algorithms that predict measurement values from acquired photos. It can read circular, semicircular, and linear gauges, digital and rolling or analog digit displays, and colored warning lights, among other things. At no additional cost, a single camera can be utilized to monitor several gauges. Computer vision and machine learning algorithms automatically extract gauge measurements from photographs, and the findings may be easily validated and corrected if necessary. The data captured can be pushed into current equipment maintenance systems using an API.

This Japanese company has recently raised $2.2 million in early 2021. LiLz began with basic building management but has since expanded to include chemical and industrial plants, construction and manufacturing sites, steel plants, public infrastructure like roads, and other areas.

The company is continuously updating its technology. They’ve mostly been refining the software, making it more remote-updateable, enhancing accuracy and robustness to disturbances, and adding data sharing and other functions. They are also developing new products and technologies for functions like sound search and counting things in the camera’s field of view, monitoring float and level gauges, etc.