Date
Monday, July 28, 2025
Time
11:00 AM - 11:30 AM
Location Name
Room 300D
Name
Using AI at Treatment Facilities to Save Time and Money
Track
Data Analytics
Description
Plant operations often rely on reactive measures, addressing issues only after equipment malfunctions or performance drops. This approach depends on lagging indicators and troubleshooting, rather than real-time or predictive analysis. The industry also faces a shortage of skilled O&M staff, leading to conservative "set it and forget it" decisions, inefficiency, and compliance risks. Overburdened field staff can make impactful decisions if equipped with predictive tools they trust. Jacobs O&M, data scientists, and Palantir developed digital tools now used at dozens of sites, offering over 50 successful use cases. These tools function like smartphone driving directions, providing recommendations to O&M staff, who retain decision-making authority. Successful technology adoption requires staff trust and engagement, as past failures showed. Tools must integrate seamlessly into daily workflows and involve staff in development and deployment. The tools were part of a 10-year digital transformation at Jacobs O&M, starting in 2018, aimed at supporting field staff with data science, machine learning, and AI. Key areas addressed include chemical dosage control, aeration control, and maintenance scheduling. Development faced challenges like cybersecurity, varied data sources, data quality, and user trust. The deployment process involves preliminary analysis, data connection, initial deployment, and ongoing support. Early engagement of field staff, focusing on solving their problems, training, and support are crucial for success. Other factors include having an onsite champion, executive support, pacing adoption, clear performance baselines, secure data handling, and using available data. Results have been excellent, with acceptance rates of 65-90% for recommendations. Chemical and power savings range from 10-30%, with perfect regulatory compliance. Maintenance tools save time but rely on staff expertise. AI tools analyzing historical data have been popular and cost-effective, with quick payback periods. The tools have also fostered discussions on best practices, enhancing digital maturity and decision-making skills.