Waste removal is quicker and safer with input from AI and digital twins

^{Lanner uses digital twins for safe and efficient waste removal}

A huge clean-up operation is under way in the United States to clear more than 56 million gallons of nuclear waste from a site to protect a nearby river and surrounding habitat. One of the challenges facing the clean-up team is the sheer complexity of decision making. A single question about how to improve throughput may involve 80 possible scenarios, taking a month to provide an answer.

Our predictive simulation experts at Lanner, working with Washington River Protection Systems, have come up with a solution. In combination with digital twins, they are using cloud computing, artificial intelligence and machine learning to support the safety and efficiency of the clean up and identify where improvements can be made through automatic bottleneck detection.

The impact has been to dramatically reduce overall computing time, so that these complex questions can be answered in a couple of days, rather than 25. The simulations provide detailed insight into how to plan next stages which has boosted efficiency and reduced human error. For example, in assessing a proposed investment to boost throughput at an effluent treatment plant, the model revealed a bottleneck in the filtration system. Without detection, it would have slowed throughput by 75% compared to targets. The work has consistently enabled challenging deadlines to be met, avoiding fines levied for late removal.