日本語 
Fracta Japan Co., Ltd. and Toho Gas Network Co., Ltd. have successfully developed an AI-based algorithm that predicts the deterioration of internal gas supply pipes, marking a first for the city gas industry. The companies have also confirmed that using the technology to prioritize pipe replacement delivers nearly double the return on investment compared to the conventional approach of replacing pipes solely based on their age.
The development is the result of ongoing collaboration between the two companies, which began as Fracta prepared to expand into the gas sector. Previously, Toho Gas Network had adopted deterioration prediction AI jointly developed with Fracta for gas mains, using the technology to improve the efficiency of replacement planning.
One of the key hurdles in underground pipeline management is that deterioration doesn’t really happen at a single steady rate. Things like soil conditions and weather patterns can really steer the pace of how quickly pipes degrade, and it’s not uniform, at least not in a way that’s easy to predict. The newly developed AI builds on operational know-how gained from gas main applications and combines it with gas pipeline data, allowing deterioration prediction capabilities to be extended to internal gas supply pipes.
Also Read: NTT and the UK Forge Frontier Technology Partnership, Strengthening Japan’s Global AI and Innovation Leadership
Toho Gas Network plans to use these predictions to support safer and more efficient replacement of supply pipes located beneath roads. The technology will also help identify aging internal pipes on customer premises and support replacement recommendations where necessary.
Looking ahead, Fracta and Toho Gas Network will continue working together to advance deterioration prediction technologies for internal gas infrastructure. The broader goal is to make the solution available to other infrastructure operators across Japan and contribute to more efficient replacement of aging pipeline networks.
English