‘AI is transforming infrastructure’ says Bentley Systems CEO Nicholas Cumins

Just two weeks after Water Environment Federation and Amazon announced the Water-AI Nexus Center of Excellence, Bentley Systems showed a vision for how AI will transform more than just water infrastructure.

During its annual Year In Infrastructure and Going Digital Awards event held in Amsterdam Oct. 15 and 16, Bentley Systems CEO Nicholas Cumins pointed to aging infrastructure, extreme weather events, and the AI boom driving data center construction globally as drivers behind this AI transformation.

"We're still in the early days,” Cumins said, “but the signals from this year's submissions are strong: AI is transforming infrastructure."   

As an example, he spoke of a leak detection system in Brazil in which AI was used to detect leaks throughout the system. Detections were reviewed by engineers and water professionals resulting in a 1/3 reduction in water outage complaints while saving 175,000 cubic meters (46 million gallons) of water per month.

AI’s implementation in Bentley Software

While using AI to identify and detect leaks is not the most novel use of AI in today’s water market, Cumins said that chatbots and agents will redefine how engineer ‘s workflows. In particular, these large language models allow engineers to use every day language to solve complex problems.

As an example, he posed a hypothetical bridge project for which an engineer needs a to create a foundation for construction. That engineer could simply ask the software, “How did we design the foundation for a similar bridge last year,” to use existing knowledge to accelerate the work of the latest project.

While that example was not specific to the water market, it does not take much to imagine using an AI agent to ask similar questions about aeration basin sizes, valve sizing and pipe diameters, treatment technologies based on similar footprints or tank depths or location for pressure management.

Consider also that workforce issues has been a persistent area of discussion throughout the industry, which is undergoing a generational shift as the Baby Boomer generation retires and hands off the reins to Generation X and Millennial professionals. This implementation of AI in Bentley Systems software is an example for how large language models could address legacy knowledge transfer.

Live demonstration about stormwater at a data center

During the keynote presentation Oct. 15, Bentley Systems executives showcased a live demonstration of the agent in action. For this example, an engineer is using OpenSite+ to better understand stormwater retention for a data center site.

The engineer first asks for the total volume capacity of the ponds on the site, which returns a response of 3,000,000 cubic meters (22.4 million gallons) for the total volume of all three ponds onsite.

He then asked it to “Calculate the runoff for this 480 acre site for a four-inch-per-hour storm and a sea value of 0.9,” which returns a 6,272,640 cubic feet or 144 acre feet. It also showed its work by revealing each step of the calculation and the formulas it used to arrive at the final result.

The whole demonstration was done using voice to text rather than typed questions, and while the engineer’s accent presented a difficulty toward the end of the live demonstration, typing in questions is still an option for exacting the language to get the desired question answered.

Read more about how Bentley is implementing AI throughout its product suite with these press releases and blogs.

• Bentley Systems advances infrastructure AI with new applications and industry collaboration

• Bentley Unveils AI-Enabled Generation of Open Applications