By Mike Franklin, Chief Information Officer, BGE, Inc.
Two hundred feet below downtown Austin’s streets, engineers made a decision that transformed one of the state’s most ambitious drainage projects.
Capital Express (CapEx), part of the Texas Department of Transportation’s (TxDOT) $4.5 billion reconstruction of Interstate 35 (I-35), shows what can happen when architecture, engineering and construction (AEC) firms commit to digital transformation. It also highlights that sustaining booming growth across the southern United States demands resilient infrastructure built smarter, faster and with less risk.
As CIO of BGE, Inc., the firm that led this design work, I’ve watched competitive advantage in our industry shift from speed-to-delivery to speed-to-insight. Advanced tools like digital twins and virtual reality are advancing speed-to-insight at a pace conventional workflows can’t match.
From Blueprints to Digital Twins
When our engineering consultants joined the CapEx project in January 2023, they began assessing how best to scale a massive underground drainage tunnel system to manage stormwater and mitigate flooding.
The original design plans called for three small tunnels, each 8 to 10 feet in diameter, placed between 30 and 40 feet underground in difficult geology. Industry standards put this approach at the edge of what was physically achievable, with a real risk of significant disruption to surrounding neighborhoods.
Computational modeling changed the outcome. By building an integrated digital model of the corridor, bringing together geotechnical data, hydraulic simulation, utility records and surface conditions into a single environment, engineers could evaluate flow capacity, tunnel alignments and pump station configurations in hours rather than weeks. Four preliminary studies were completed in 3 months, unlike typical processes that can take 12 months.
The tunnel system that emerged bore little resemblance to the original concept. Spanning 6 miles of 22-foot-diameter tunnels, positioned nearly 200 feet below ground in the Austin Chalk formation, one of the world’s most stable tunneling mediums. Dropping into stable geology eliminated the need for specialized excavation equipment, extensive ground stabilization and the cascading construction risks that follow.
The estimated savings exceeded $200 million, unlocked not by luck, but by better data that enabled faster, smarter decisions.
The Data Infrastructure Behind the Engineering
Designing a 6-mile tunnel system beneath a dense urban environment means navigating neighborhoods, historic structures, active roadways and existing utilities simultaneously. Aligning dozens of engineers, agency officials and community stakeholders on decisions depended on solving a data infrastructure problem: shared visibility.
Over 18 months, the CapEx team facilitated more than 1,000 meetings, engaging 175+ professionals across numerous departments and agencies. What made that pace sustainable wasn’t any single design tool. Instead, it was the common data environment underneath: standardized platforms that let every discipline work from the same integrated model.
Large-scale projects like this are increasingly deploying 360-degree, high-resolution camera systems, letting engineers walk the tunnel alignment digitally and conduct virtual site visits with remote stakeholders in real time. Integrating data into the same shared model ensures virtual walkthroughs and design reviews look at a single source of truth. The results: fewer costly field visits, reduced safety exposure and faster decisions.
The hardest part of this transformation isn’t the technology. It’s changing habits and convincing everyone to trust a shared model over the siloed approaches and processes they have relied on for decades. The technology investment only pays off when the organization around it changes, too.
Simulation Before First Shovel
The Cesar Chavez Pump Station, the mechanical heart of the CapEx system, was designed around four concrete volute pumps that discharge I-35 stormwater runoff directly into the Colorado River. The pumps are a proven worldwide solution. What’s changed is how they are selected and sized, using computational fluid dynamics (CFD) modeling to simulate full-scale hydraulic performance before construction begins.
Engineers also built the pump station above the 500-year floodplain, protecting pump motors valued at over $100 million from catastrophic weather events. The system is designed to operate through a 100-year flood. Once it exceeds that threshold, it shuts down safely to protect equipment and can be reactivated immediately after the storm passes. Without that important design choice, recovery could take months. In worst-case scenarios, years.
Significantly, no other public infrastructure system of this scale in Texas incorporates that standard. That’s why scenario modeling is a step change, opening up a realm of new possibilities as data challenges previously held conventions.
Where AI Fits and Where It Doesn’t
The AEC industry is in the midst of an AI transformation, where firms must balance ambition with healthy skepticism. On projects like CapEx, machine learning delivers near-term value by accelerating workflows rather than delivering core tunnel design. Key areas of impact include unglamorous data tasks, supporting human-in-the-loop validation and strategic risk reduction.
On multi-billion-dollar projects, advanced models and AI don’t replace an expert’s judgment. Rather, they extend its reach, helping us validate the right decision at the right time.
Practical examples of the AI impact include its ability to undertake automated document reviews, parse complex regulatory codes and organize unstructured site logs. It can also run rapid simulations to identify project bottlenecks and resource clashes before ground breaks and cross-reference historical line items, flagging budget variances faster than manual audits.
Supporting the human in the loop, AI can run algorithmic checks to test assumptions and confirm high-stakes choices under tight deadlines. It can also help mitigate project risk by taking on repetitive tasks, reducing the chance of human error and letting experts focus on complex physical problem-solving.
Building For What Comes Next
The mega-projects of the next decade will be shaped by climate resilience demands, urban growth and aging infrastructure. Closing the gap between an original design concept and final execution shows how advanced simulation and visualization technologies are redefining and revolutionizing what’s possible in the modern built environment.
For technology leaders, the mandate is clear. Project success now requires data infrastructure that turns data into actionable insight for faster decision-making.
Two hundred feet below Austin, that gap was closed in just 90 days, saving millions—a result that will serve the city for generations.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
