NTT DOCOMO, NTT Communications and NTT Comware announced today that they will begin providing a trial version of an AI-based system for monitoring the water levels of rivers and reservoirs to companies and local governments nationwide, effective immediately.
Through the trial, the companies expect to improve the system's AI functions and usability, aiming at a commercial launch by March 2024.
Notably, the system eliminates the need for physical water-level gauges and instead uses virtual gauges and AI technology to visualize real-time images and time-series data of water levels on a screen. Also, the system can automatically notify authorities if a predetermined water level is exceeded.
Basic system
The system uses cameras and other dedicated equipment installed near a river or reservoir to capture video images of the water. A virtual gauge is overlaid on the video and AI video-recognition technology accurately determines the water level in real time. Specifically, the system uses a unique algorithm and AI image-segmentation technology to analyze the ratio of the water surface to the land area displayed on the virtual gauge. Accuracy can be improved by using multiple virtual gauges.
Virtual water-level gauge set up at four points on a screen
Personnel in charge of monitoring rivers and reservoirs for municipalities, etc., can check water levels and configure AI settings using video and graphs on a management screen in the cloud. In addition, the system can be live-streamed to local residents to keep them alert and informed.
In addition to the technology for determining water levels, the system incorporates two other key technologies:
EDGEMATRIX®
NTT Com's video analytics platform (video edge AI platform) processes high-volume video data in an Edge AI Box (edge computer), enabling rapid video analysis, water-level determination and streaming.
SmartMainTech®
DOCOMO, as part of its SmartMainTech series of solutions for infrastructure maintenance, has developed an AI for determining water levels (joint patent pending for DOCOMO and NTT Comware), which combines DOCOMO's AI technology and a video-analysis AI in Edge AI Box, to analyze video of rivers and reservoirs and link the results to the SmartMainTech cloud. The solution incorporates industry-specific AI, data-analysis and visualization technologies.
The system has many noteworthy features:
1. Use of virtual water-level gauges
Low-cost virtual gauges make it practical to monitor even small rivers and reservoirs.
Virtual gauges cannot be damaged by floods, etc., and are much easier to maintain than real gauges.
Multiple virtual gauges can be installed for each camera to increase system accuracy.
2. Use of real-time video
Rapid water level changes can be confirmed via real-time video.
3. Batch video and data monitoring
Real-time video and water level graphs can be viewed on the same management screen.
Notifications can be sent automatically via email or text if a specified water level is exceeded.
4. Easy integration with video services
Video can be easily linked to video-distribution sites to keep nearby residents up to date.
Edge AI Box performs video encoding in the field, enabling real-time video delivery.
Each company performed a specific role in the development of the system.
DOCOMO: Developed the AI used in SmartMainTech
NTT Com: Provided EDGEMATRIX service and cameras and Edge AI Box
NTT Comware: Developed AI applications under the SmartMainTech brand
Through the trial, the three companies expect to confirm and optimize the system's deployment and accuracy, with a view to launching it on a commercial basis for municipalities nationwide as soon as possible.
In recent years, river floods caused by typhoons and torrential rains have become more severe due to climate change, increasing the need to accurately ascertain water levels and take proper countermeasures quickly. However, the use of conventional physical water-level gauges is costly and labor-intensive in terms of installation and maintenance. Going forward, the DOCOMO group's low-cost, low-maintenance system is expected to expand the vital real-time monitoring of not only large bodies of water, but also small rivers and reservoirs that have the potential to cause disasters.
