Going Digital 2023: Towards Infrastructure Intelligence

It’s Singapore. You have to love the infrastructure.

Bentley Systems has sponsored this post.

The 2023 Year in Infrastructure and Going Digital Awards event, Bentley Systems’ annual awards ceremony in which the grandest infrastructure projects are honored, was held at Marina Bay Sands.

“A warm welcome to Singapore,” says Bentley SVP, Kaushik Chakraborty, Regional Executive of the APAC region, and he’s not joking. Outside the air is thick and steamy, with temperatures expected to be in the 90s (⁰F) all week.

(Image: Bentley.)

(Image: Bentley.)

“The weather gods were favorable this week,” Chakraborty adds. A Singapore native, he appreciates the dip into the high 80s, as during the previous night, when Bentley reserved the observation deck of the Marina Bay Sands hotel for its reception.

The Marina Bay Sands is a story in itself. An iconic 57-story structure on which floats a gigantic infinity pool, it was the winner of Bentley’s annual award in 2019, when they were called Be Inspired.

The Marina Bay Sands Hotel towers above the convention center in Singapore Marina Bay, site of the 2023 Year in Infrastructure and Going Digital Awards. (Image: Bentley.)

The Marina Bay Sands Hotel towers above the convention center in Singapore Marina Bay, site of the 2023 Year in Infrastructure and Going Digital Awards. (Image: Bentley.)

“This is probably one of the greenest cities in the world,” he continues. When asked why he likes living in Singapore, he tells people it’s because of the fantastic infrastructure. He is most proud of the SMRT subway. “We hear from visitors who come here from New York, and when they ride the subway, they’re completely dazzled at how effective and how clean it is.”

Enter Greg Bentley, CEO

Singapore can count on a fan even higher placed at Bentley Systems. Bentley Systems’ CEO Greg Bentley has identified Singapore as a shining example of technology in the Eastern Hemisphere. China may have more large infrastructure projects, but Singapore leads in technology implementation per capita.

Bentley’s Year in Infrastructure alternates between Singapore and Europe, lately mostly in London, and the nominees reflect the venue. This year there were 302 nominees from 51 countries, with the majority from Asia Pacific and the rest from Europe and the United States.

Greg refers to a survey of infrastructure engineering firms, which reveals that they have an average backlog of a year. He thinks this is making staff engineers work an extra 23 minutes more each day compared to pre-pandemic years.

Greg Bentley, CEO, Bentley Systems. (Image courtesy of Bentley Systems.)

Greg Bentley, CEO, Bentley Systems. (Image courtesy of Bentley Systems.)

Greg uses the YII event to promote the company’s user projects. This year, Bentley Systems required each project to reveal the savings resulting from going digital. The results showed an average savings of 18 percent, but there were process improvements, as well, such as automation, replication, standardization and technology adoption. Specifically, 28 percent of submissions used iTwin Capture and 17 percent used SYNCHRO for reality modeling and 4D construction modeling.

(Image courtesy of Bentley Systems.)

(Image courtesy of Bentley Systems.)

In addition to those nominated that undergo independent judging, Bentley Systems picks additional projects to include. These projects may be outside the categories of the independently judged projects. Those selected in this manner are awarded “Founders’ Honors.”

Greg refers to this New York Times article about Singapore’s plan to forestall rising atmospheric temperatures. Already sweltering, the city cannot get any hotter without endangering the lives of its inhabitants—such as in places like “Air Conditioner Alley,” a narrow street lined with rowhomes all with window units that dump heat into the street, making things worse. Singapore’s solution: more trees, heat absorbing (not reflecting) paint and orientation of buildings for natural cooling effects, among other things.

Greg Bentley, 3rd from the left in the picture above, was included in a U.S. Department of Transportation meeting led by U.S. Transportation Secretary Buttigieg. Bentley Systems counts the U.S. government as a special customer, which has had a perennial #1 spot on Bentley’s Infrastructure 500 Top Owners.

Greg made a case for going digital to support the U.S. government’s Infrastructure Investment and Jobs Act (IIJA), where a lack of skilled labor is delaying some projects. Bentley also proposed a smarter type of maintenance and used Singapore’s rail system, SMRT, as an example.

“In Singapore, our AssetWise Linear Analytics solution is the basis for transit agency SMRT’s Predictive Decision Support System, optimizing so that only effective and necessary maintenance is done at the right time, in place, while improving reliability to the target of at least 1 million kilometers between five-minute delays,” says Greg.

Singapore’s system relies on sensor data that allows comparison to the digital twin of the SMRT tracks to determine conditions that warrant actual maintenance, rather than a set schedule that uses staff resources for maintenance whether needed or not.

Bentley has been emphasizing digital twins since last year’s Year in Infrastructure and customers seem to have caught on. Bentley points to the increasing rate of components being added in ProjectWise, Bentley’s asset management application—now over 100 million a month.

Two thirds of this year’s Year in Infrastructure projects credit Bentley’s iTwin Platform.

Another way AI is being used in Singapore is its Public Utilities Board applying AI in OpenFlow’s Darwin genetic algorithms and deep learning from pressure and volume sensor data to predict where and when leaks would occur. Parallel processing of acoustic signal analysis allows the system to work in near real time, says Greg.

Generative AI

Keith Bentley, Greg’s brother and former Bentley Systems CTO, missed this Year in Infrastructure. He was home with a fever – the AI fever that started with ChatGPT, which has infected every conference. Specifically, Keith has been studying Copilot, Microsoft and GitHub’s ChatGPT-based assistant to developers.

Greg concedes that software engineering isn’t the same as infrastructure engineering but shares many of its aspects. A veteran developer, Keith finds more time for creative coding, leaving mundane development, documentation and testing to Copilot. His goal: bringing something like Copilot for code to civil, structural and geotechnical engineers for design and simulation.

Greg finds Copilot’s use of all sundry available source code a little too democratic. In all that code is code that is less than optimum. Managers of coding talent, like Keith, know how to distinguish good code from the rest just like engineers and architect managers can distinguish good design. A lot of that good design from Bentley users is on the cloud, but the question is how to train AI on good user design without exposing users’ IP to other users. Or as Greg puts it, “an involuntary contribution for everyone else’s unwitting misappropriation.”

“Together, we can make sure their [users’] accumulated digital components are sufficiently robust and capable in their own right for generative training in their specific domains without presumptuous commingling of intellectual property,” says Greg.

While this approach limits the size of the training database, it fits what Greg refers to as being good stewards of their users’ data.

“Our accounts, not Bentley Systems, own their data on Bentley Infrastructure Cloud. Generative AI doesn’t change our stewardship commitment,” says Greg.

Bentley uses Mott MacDonald, a U.K.-based AEC firm, as an example of porting components used on a water utility project into ProjectWise Component Center. The company was at the conference as a Going Digital Award finalist.

Australia-based Laing O’Rourke train station project with SYNCHRO 4D modeling provided many components that can be reused for future projects, allowing for “modular reuse at full scale,” says Bentley.

Another Bentley user, KBR used generative components to design and model just once, then proceed to the excavation stage benching plan letting the various and numerous footings across the elevated train line be automatically configured.

Hyundai Engineering used its 1,680 plant structure designs in ProjectWise, applied machine learning to classify their purposes and capacities, and now can recommend the most appropriate project that can be modified for next projects.

To learn more about the Going Digital Awards in Infrastructure, visit Bentley.