As utilities face rising infrastructure costs, aging assets and growing pressure to modernize essential services, artificial intelligence is finding a new role in engineering design. But unlike applications that rely on AI to generate recommendations, a new approach seeks to combine the flexibility of conversational AI with the precision of established engineering calculations.

Princeton, New Jersey-based Transcend has introduced two AI capabilities for its Transcend Design Generator (TDG), a platform used to develop preliminary designs for water, wastewater and power infrastructure. Announced October 13, the additions aim to help utilities, engineering firms and equipment manufacturers evaluate infrastructure investments more efficiently while keeping technical calculations grounded in established engineering principles.

Connecting AI to Verifiable Engineering

The new capabilities, Ask TDG and MCP Gateway, address different aspects of the infrastructure planning process. Ask TDG provides a conversational assistant within the platform, allowing engineers and project stakeholders to ask questions about design methodologies, calculations and equipment selection. Its responses draw from TDG’s technical documentation, helping users interpret engineering results without altering the underlying calculations.

MCP Gateway extends those capabilities into enterprise AI environments through the Model Context Protocol, an open standard that allows AI systems to interact with external tools and data. Authorized users can connect their existing AI applications to selected TDG functions, allowing engineering outputs to support project analytics, capital planning and downstream workflows. Importantly, the engineering calculations remain within TDG rather than being generated by a probabilistic AI model.

This distinction addresses a significant concern surrounding AI adoption in critical infrastructure: the risk of inaccurate or fabricated technical information. TDG relies on deterministic, first-principles engineering logic, meaning calculations follow defined rules and can be traced to established engineering methods and utility-specific standards. Although this architecture may reduce certain risks associated with AI-generated designs, it does not eliminate the need for engineering verification, particularly when projects involve complex site conditions, regulatory requirements or long-term operational risks.

Bringing Engineering Into Capital Planning Earlier

Traditional infrastructure planning often begins with preliminary cost assumptions that become more detailed as engineering progresses. For utilities managing extensive capital improvement programs, that process can create uncertainty around project budgets, equipment requirements and construction schedules. Transcend’s approach is designed to introduce more detailed engineering information before investment decisions are finalized.

TDG allows users to enter project parameters such as treatment capacity, electrical loads, site conditions and operational objectives. The platform then generates preliminary designs for facilities including drinking water and wastewater treatment plants, pump stations and electrical substations. According to Transcend, these designs can be produced in minutes and incorporate process diagrams, three-dimensional building information models, equipment specifications, material quantities and operating expense estimates.

Because multiple engineering disciplines operate within the same design environment, adjustments to project assumptions can be reflected across mechanical, electrical, civil, structural and process requirements. For capital planners, this could make it easier to compare alternative technologies, evaluate lifecycle costs and identify potential budget constraints before committing to a preferred design. However, the financial value of faster preliminary engineering will depend on how accurately those early designs translate into detailed project specifications and eventual construction costs.

A Broader Role for AI in Utility Operations

Transcend is also building connections between utility design standards and equipment manufacturers through TDG Nexus, an existing system for sharing approved technical specifications. Utilities can incorporate their own engineering requirements, while technology providers such as Xylem and DuPont can make equipment specifications available for use in preliminary designs. The combination could reduce repetitive specification work and improve consistency between early planning and subsequent engineering stages.

The company reports that TDG has been used to generate more than 20,000 designs across more than 120 countries, with projects collectively affecting an estimated 652 million people. Its customer base includes major water utilities such as Sabesp, Severn Trent, Scottish Water and Mekorot, along with engineering firms including WSP, STV and CHA. These figures indicate an established international presence, although Transcend has not publicly quantified the incremental cost savings or productivity improvements associated specifically with the newly introduced AI capabilities.

For utility executives and engineering organizations, the larger opportunity may lie in connecting increasingly sophisticated AI workflows with engineering systems that provide consistent, auditable results. As infrastructure owners seek to stretch limited engineering resources and make more defensible capital investments, platforms that combine accessible AI interfaces with traceable technical calculations could become valuable planning tools. The longer-term test will be whether these integrations produce measurable improvements in project delivery, cost predictability and asset performance without compromising engineering oversight.