5YF Episode #27: Gecko Robotics Co-Founder Troy Demmer

Future of Industrial Robots: A Built World Intelligence Layer

subscribe and listen today

Today, we dive into how robots are revolutionizing the industrial sector and keeping our critical infrastructure safe.

AI models today are an amalgamation of the data on the internet but there is a lot of data held in corporations that hasn’t been gathered yet.

The built world—from bridges to power plants—is undergoing a digital revolution, driven by data and AI layered onto physical assets. Gecko Robotics leads this charge, monitoring over 500,000 assets with sensor-rich robots that crawl, swim, and fly to capture game-changing insights. With $5 trillion in U.S. infrastructure urgently needing repair, Gecko’s data models are unlocking a future of predictive maintenance and autonomous management.

Troy Demmer, Co-founder and Chief Product Officer, envisions a future where AI not only maintains but transforms infrastructure, enabling industries to design and operate assets with unprecedented agility and intelligence—shaping a smarter, more resilient world.

My 5 Year Outlook:

• Autonomous Infrastructure Management: ‍Data rich asset models and digital twins across physical infrastructure.‍
• $5T Infrastructure Design Upgrade Using Software: ‍Data-led improvements from new materials to completely novel designs.‍
• AI Models Advance To Replace The Need For Data Collection: After reaching a certain scale of data and prediction quality, AI will simulate with unmatched accuracy.

Curious? Read on as I unpack each below 👇🏼

Autonomous Infrastructure Management

In the near future, our infrastructure will be managed more autonomously, blending robotics, digital twins, and AI to make data-driven decisions in real time. Every part of our built environment—from bridges and tunnels to power plants and pipelines—will have a digital twin. These virtual models allow real-time monitoring and diagnostics, capturing every detail of physical assets.

For example, Gecko Robotics is pioneering ultra-detailed digital twins of complex infrastructure, fusing data from multiple sensors from its robotic inspections. Imagine a bridge that, instead of undergoing sporadic manual checks, continuously updates its digital twin with data on wear, strain, and potential weak points. Drones scan its exterior while robots crawl over its frame complimented by embedded IoT sensors. In this future, roadways could predict and mitigate traffic wear based on vehicle loads, while water systems could adjust for peak usage, all autonomously.

This shift transforms physical infrastructure into a living, adaptable ecosystem powered by software. A digital layer that makes proactive maintenance feasible, preventing failures before they even surface. As we scale these data models across cities and states, digital twins will become essential tools for infrastructure management, fundamentally transforming how we maintain and improve our physical world.

Troy Demmer, Co-Founder and Chief Product Officer of Gecko Robotics

Gecko Robotics uses a fleet of advanced robots and AI software to help government and heavy industry maintain and manage their critical infrastructure. From navy vessels to power plants to dams, Gecko collects data on and delivers insights across over 500,000 of the worlds most important and critical infrastructure. It has developed a digital layer of intelligence over the built world to improve performance, prevent breakage and failures, and increasingly predict how an asset will behave in the future. The company has raised over $220M from top investors including Founders Fund and US Innovative Technology Fund.

Gecko’s co-founder and Chief Product Officer is Troy Demmer. A graduate from Carnegie Mellon University, Troy was previously in the healthcare industry working at the University of Pittsburgh Medical Center before launching his first startup 360Showings, which 3D rendered homes for the real estate market. Along with his work at Gecko Troy also runs his own venture firm, First Order Fund, which invests in early stage startups building moats using data.

$5T Infrastructure Design Upgrade Using Software

Increased monitoring and data-collection of assets unlocks an ability to not just maintain asset health but improve its design and function. The most important evolution of this process will be the speed and timeframes to design upgrades. Previously, assets form water tanks to pumps to ships were manually observed over years if not decades before engineers suggested improvements. The future points to a collapsing of that timeframe and effort thanks to a constantly flow of data, digitization of the asset in software, and the simulation of different upgrades using AI.

Over the next five years, software-driven insights will transform infrastructure design, creating a continuous improvement loop where data leads directly to innovations in materials and architecture. As companies like Gecko Robotics feed massive amounts of real-world data into digital models, we’ll uncover patterns that inform smarter designs and longer-lasting materials.

For instance, a water tank digital twin could analyze stress distribution and environmental wear, revealing that a specific composite material, after countless simulations of variations, outperforms traditional steel in durability and cost-effectiveness. This isn’t hypothetical—real-world data can already indicate where to reduce material use without sacrificing safety, or how to design structures that better withstand natural disasters.

We’re moving toward an era where infrastructure not only responds to current conditions but also anticipates future needs. By making every infrastructure project an evolving prototype, data-led design will ensure that the next generation of infrastructure is more resilient, adaptive, and sustainable.

AI Models Advance To Depart From The Need For Data Collection

The model may know even better than observations made in person because you have this corpus of data that has gotten so good at predicting what's next.

In five years, AI models will have progressed to a point where they can operate beyond the need for extensive, on-the-ground data collection. Trained on an enormous corpus of historical and real-time data, these models will become sophisticated enough to predict asset conditions, risks, and even maintenance needs with greater accuracy than human observation.

Imagine an AI model trained on the behavior of thousands of miles of pipeline. By recognizing patterns in temperature fluctuations, pressure changes, and material fatigue across assets, it can project the performance of future builds and pinpoint failure risks in existing infrastructure—often more reliably than if an inspector were physically present.

For instance, Gecko Robotics has data models covering over 500,000 assets. If they’ve gathered data on 1,000 identical water tanks, their model will be able to predict with precision how the 1,001st tank will perform. This level of predictability means that the need for fresh data collection diminishes as the dataset scales and model accuracy grows. Instead of continuous inspections, Gecko’s model leverages its vast, high-fidelity dataset to monitor new tanks remotely, setting a new standard for proactive asset management.

This shift doesn’t just improve efficiency; it redefines what’s possible. Inspections for remote, high-risk areas like isolated pipelines and offshore rigs can now be simulated by advanced AI models, enabling continuous monitoring in locations where human access is challenging or costly.

In effect, AI will provide a sharper, more consistent lens on the built world, transforming how we monitor, repair, and even design resilient infrastructure for the future.

Let’s build!