Situation: Operations professionals at utility, industrial and commercial sites face intense pressure to maintain high levels of performance and reliability while controlling the cost of operations and maintenance. nDimensional’s client, a global leader in engineering, procurement and construction services, offers Monitoring & Diagnostics (M&D) services and software to help plant operations professionals shorten issue response times, respond to problems well before they trigger alarms or force shutdowns, and extend the life of infrastructure assets.
Solution: nD was used to create digital twins with monitoring and diagnostics that learn healthy operating patterns, auto-identify anomalies and streamline diagnosis in conjunction with the client’s in-house software system and domain expertise. For each M&D customer, asset and system-level digital twins were created and connected to myriad sensor, device, and subsystem data. Hybrid engineering and machine learning models were used to benchmark healthy asset behavior patterns, alert operations personnel of anomalies and streamline problem isolation and diagnosis. nD continuously processed 100,000’s of near-real time data streams, 10,000’s of machine learning models and 1000’s of engineering models, with automated ML model retuning and engineering model advanced version management. The solution helped our client’s M&D center reduce false positives by 70% and helped our client’s customers to realize over $150M in risk-weighted savings through the early detection and remediation of asset performance and health issues.