A digital model helped engineers redesign a major drainage system beneath Austin, with the project’s engineering firm estimating more than $200 million in savings. The case shows how shared data and faster modelling can change infrastructure decisions, while AI’s role in the article is broader than this particular project.
Watch Desk analysis
What happened
Michael Franklin, CIO of engineering firm BGE, says the team working on Texas’s I-35 Capital Express project brought geological, hydraulic, utility and surface data into one model. Four preliminary studies were completed in three months, compared with the year he says such work would typically take.
The resulting plan called for six miles of 22-foot-wide tunnels nearly 200 feet underground, rather than three smaller tunnels. Franklin says choosing a route through stable geology avoided specialist excavation and ground-stabilisation work, contributing to estimated savings above $200 million. He also describes shared project data and virtual site visits as ways to coordinate the many teams involved. Read Franklin’s account in Forbes.
The article also discusses AI uses in the architecture, engineering and construction industry, including document review, regulatory-code analysis and organising site logs. It does not establish that AI produced the Capital Express design or its reported savings.
Why it matters
The project is a useful example of computing tools influencing decisions before construction begins. Combining data in one model helped engineers compare options more quickly; the reported savings came from the resulting design choices, not from a claim that AI designed the tunnels.
Franklin writes as the CIO of the firm that led the design work, so the savings and comparisons are his account of the project. The practical takeaway is less “let the algorithm build a tunnel” and more that joined-up data can help specialists test consequential choices earlier.
Our read
There is real substance here, provided we keep the tools straight. Digital modelling and shared data are the demonstrated centre of this case; AI appears as a wider set of possible industry applications. That distinction matters, even if it is less glamorous than giving every clever spreadsheet a robot’s hat.
For infrastructure teams, the example makes a case for investing in data that different disciplines can actually use together. The next question is whether the claimed speed and savings can be shown in other projects, not just described in one impressive case study.
What to watch
- Whether the project’s reported savings are documented beyond Franklin’s account.
- Whether other infrastructure projects report comparable results from integrated modelling.
- Which AI applications are actually deployed in engineering work, and with what measured effects.
Discussion spark: For major infrastructure projects, should the priority be a shared digital model that helps experts compare options, or more direct use of AI to automate parts of the design work?
Sources and evidence
- Why Speed-To-Insight Is Redefining Mega-Infrastructure Projects – Forbes (7 October 2026, 14:15 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.