We like working on interesting challenges. We love solving them 😊. Reading analog instruments in a manufacturing setting is one such challenge that we were focused on recently.
While analog gauges have long been the backbone of many industrial processes, many of them are still manually read and measurements transcribed from notes taken on the factory grounds – cumbersome and error-prone tasks. That’s where we stepped in and applied our ML/computer vision skills – using advanced object detection and instance segmentation, we build a solution that can quickly and accurately read analog gauges, saving time and reducing errors.
With our solution, field operators no longer have to record data manually. Instead, the solution detects gauges automatically, reads their current measurement values, and transfers all readings to a central database, giving them easy access to accurate data that can be analyzed and acted upon.
The solution was developed in response to the TechEngine call of NIS.