The newly-developed AI software enables highly accurate detection of underground objects using ground-penetrating radar (GPR).
September 29, 2021 – Kyoto, Japan – HACARUS INC, the leading provider of big insights from small data, today announced the launch of a new AI software module, the result of a successful joint development project with Osaka Gas Co., Ltd.. The new software can be used to accurately locate underground objects such as gas and water pipes, communication cables, and electrical lines in ground-penetrating data. Osaka Gas plans to commercialize the software through NIPPON SIGNAL CO., LTD., the leading provider of transportation infrastructure in Japan.
Example of ground-penetrating radar usage
Before the joint development project with HACARUS, Osaka Gas had relied on specially trained workers to read and analyze ground-penetrating radar images. With HACARUS AI, it has now become possible for anyone to run underground object analysis with a detection rate of 89% – 10% higher than human-run detection methods.
Ground-penetrating radar images, before and after AI analysis
The newly-available software utilizes HACARUS’ proprietary AI technology that provides powerful yet energy-efficient solutions on edge devices, without any need for cloud communication or big data. By digitizing the knowledge and know-how of skilled workers, HACARUS aims to address the issue of skilled workers shortage – a pressing issue for the manufacturing and infrastructure industries in light of the aging population of Japan.
About HACARUS
HACARUS INC. provides big insights from small data and has since its founding in 2014 supplied solutions in 100+ AI projects across the Medical and Manufacturing fields. Headquartered in Kyoto, Japan, and backed by Osaka Gas and Miyako Capital (Kyoto University) among others, its technology enables humans to make better, faster, and more reliable decisions, based on data-driven insights. HACARUS’ proprietary AI engine is built using Sparse Modeling, a method that understands data like a human would – by its unique key features and is far more resource, time, and energy-efficient when compared to Deep Learning. To learn more, visit https://hacarus.com
