Building Multiple Levels of Trust through Technology

Trust has become one of the most valuable forms of infrastructure for modern utilities. Customers, regulators, and elected officials want confidence that utilities are investing wisely, operating transparently, and addressing affordability and equity in measurable ways. The Portland Water Bureau (PWB) and the Bureau of Environmental Services (BES) are advancing this work through the careful use of artificial intelligence, data analytics, and equity-driven planning tools. Portland's approach recognizes that building trust begins long before technology is fully implemented. Both bureaus are using AI and data to inform system investments, improve operational planning, and design affordability programs that reach those most in need. System-Level Trust – Data-Driven Capital and Asset Decisions Both utilities are applying optimization and machine-learning tools to improve transparency and efficiency in large, complex investment decisions. At PWB, the Engineering Planning group has used Optimatics AssetAdvanced since 2021 to support in-house mains project selection. The software uses the bureau's GIS pipe data, RANK database, and asset-management metrics to identify roughly 30,000 feet of the most desirable pipe segments for replacement each year. In FY 2024/ 2025, full in-house selection was completed using the software, supported by engineering review, saving extensive analysis time compared with manual methods. The process continues for FY 2025/ 2026 and will improve as data quality and internal coordination strengthen. PWB also piloted AssetAdvanced's machine-learning likelihood-of-failure (LOF) module. Early results showed high accuracy relative to statistical models, demonstrating potential to replace more labor-intensive analysis. Whether the module becomes a budgeted tool will depend on fiscal capacity, but the pilot confirmed that data-driven risk modeling can enhance affordability and accountability. At BES, staff have used Optimatics Optimizer since 2019 to prioritize capital improvements, evaluate alternatives, and inform operational strategies for the combined-sewer system. Optimizer's genetic algorithms link directly with hydraulic models and risk data to evaluate thousands of options under real-world constraints such as cost, surcharge potential, and combined-sewer-overflow (CSO) risk. Applications include: Balancing inflow control and pipe upgrades: In the Sullivan Basin, Optimizer tested combinations of stormwater controls and pipe replacements to reduce surcharge and capacity issues. Streamlined alternatives development: Automated analyses now generate and compare preliminary alternatives systematically, improving consistency and saving staff time. Large-diameter pipe evaluation: BES is piloting the Optimizer Blockage Tool to assess failure consequences for large-diameter sewers, supporting climate-resilient, long-term planning. Hydraulic-model efficiency: Work with consultants is exploring Optimizer's role in model calibration to increase accuracy and reduce manual workload. Together, these efforts demonstrate how AI can replace traditional, resource-intensive methods with scalable, transparent, and reproducible processes that strengthen fiscal stewardship. Service-Level Trust – Laying the Groundwork for Transparency At the service level, both bureaus are preparing for Advanced Metering Infrastructure (AMI). They are in the procurement phase for a meter-data-management (MDM) system and have begun installing and retrofitting AMI-ready meters. These early steps are guided by principles of transparency, privacy, and equitable access. AMI will eventually provide customers with more predictable bills, real-time leak detection, and clearer understanding of usage. Even in procurement, the conversation around data standards and integration is improving internal communication and setting expectations for accountable implementation. Personalized Trust – Engaging Customers Through Data and Design To extend transparency to the individual level, the bureaus are developing a Customer Engagement Portal (CEP) scheduled for launch next year. The portal will allow customers to view usage data, receive tailored insights, and provide feedback, creating a two-way exchange between the utility and its customers. This capability will help transform raw data into understanding and understanding into trust-well before all technical components are complete. Machine Learning for Affordability – The Smart Discount Pilot Portland is also testing the role of AI in proactive affordability assistance. The Smart Discount Pilot uses machine-learning models to analyze billing and payment patterns alongside other data indicators to identify households that may experience financial stress. The goal is to determine whether data can responsibly predict financial need and enable targeted assistance before accounts become delinquent. Early findings are promising and emphasize the importance of coupling technology with empathy, ethics, and careful governance to avoid bias. Equity and Leadership Integration Both bureaus are using an Equity Data Toolkit-which integrates utility, billing, and public-health datasets-to help prioritize infrastructure investments and program outreach. While the bureaus do not yet operate under a formal shared data-governance structure, collaboration between planning, asset-management, and customer-service teams has increased alignment and built a stronger foundation for future governance. Leadership across both organizations is emphasizing transparency, communication, and equity considerations as part of every technology-related decision. Lessons Learned and Broader Applicability Although much of this work is still underway, several lessons have already emerged: Cross-functional collaboration accelerates progress and reduces redundancy between bureaus. Early attention to data integration and privacy prevents larger issues later in implementation. Equity-driven analytics make investment decisions more defensible to both policymakers and the public. Incremental deployment allows utilities to learn and adapt without undermining public confidence. Together, these projects show that AI and analytics can strengthen trust long before outcomes are fully realized. Portland's experience demonstrates that transparency in the process-how data are used, communicated, and tested-can itself build credibility and affordability. Learning Objectives 1. Explain how AI and analytics can improve capital investment planning, asset management, and affordability-program design. 2. Describe how early-stage AMI and customer-platform planning support transparency and readiness. 3. Identify leadership, equity, and communication practices that connect digital transformation to public trust.