Australia’s growing infrastructure pipeline is putting pressure on the engineering workforce to use technology to help deliver more with the people it has.

According to Vivek Kumar, Senior Regional Director for Australia and New Zealand at Bentley Systems, Australia’s infrastructure pipeline is creating a productivity challenge that cannot be solved by simply adding more people to the workforce.

Infrastructure Australia has forecast that the sector could face a shortfall of up to 300,000 workers by 2027, as governments progress major infrastructure, energy and housing projects.

Kumar says that while migration, education and training will all have a role to play to fill the gap, the scale and speed of the challenge means the industry also needs to find ways to increase the output of its existing workforce.

“The shortage is already creating pressure across the industry,” Kumar says.

He says the challenge is not simply finding enough people to fill vacant positions. If projects are delayed, rushed or delivered without sufficient resources, the consequences can flow through to project costs, timelines and ultimately the broader economy.

This is where technology will have an increasingly important role.

Getting more from existing teams

Increasing the supply of engineers remains important, but Australia cannot train and migrate its way out of the immediate gap, says Kumar.

Even attracting experienced engineers from overseas takes time, with professionals needing to become familiar with Australian standards, codes and other local requirements.

The technology response, therefore, is not about replacing engineers, it’s about giving them more capacity.

“We’re not talking about replacing human engineers with technology,” Kumar says. “It’s about augmenting the human.”

Artificial intelligence (AI) is accelerating this opportunity. Bentley, a global software company that provides computer-aided design and digital twin technology for infrastructure engineering, construction and operations, has introduced AI-enabled capabilities across several of its products, including MicroStation, ProjectWise, Synchro and Site Design.

Kumar says early feedback from customers indicates efficiency gains in the range of seven to ten times for some tasks.

“If something was taking 100 hours to do and is now taking 13 or 14 hours to do, that’s a huge gain in efficiency,” he says.

He says that even a more conservative threefold increase in engineering output could have a significant impact when applied across a workforce facing a six-figure shortfall.

The opportunity is particularly relevant to the repetitive and administrative tasks that consume engineering time.

“Removing those tasks does not just increase productivity, it can make engineering roles more attractive by allowing professionals to spend more time applying their expertise and therefore, they are more engaged and more productive,” Kumar says.

That engagement could also help the industry attract and retain talent, particularly as experienced engineers leave the workforce and a new generation enters it.

Tackling rework at its source

Repetitive administrative tasks are only one part of the productivity equation. Kumar identifies rework as another major source of wasted time across complex infrastructure projects.

Large projects bring together multiple organisations, disciplines, software platforms and file formats. Design changes can require information to be exported, modified and checked across different systems, creating opportunities for errors and duplication.

“You have a lot of, ‘Oh no, that’s not right, I need to go back and fix that’,” Kumar says.

Connected digital environments can reduce this friction by keeping project information accessible and synchronised as designs change. Kumar points to Bentley customers using live digital twins, with design updates automatically synchronised rather than requiring teams to repeatedly return to source files.

He says one Melbourne project achieved a reported 75 per cent reduction in total rework hours, with savings of about $150 million in rework.

“That is massive,” Kumar says. “By creating a more efficient workflow, we can massively reduce the productivity issue.”

Bentley’s approach also emphasises interoperability and openness rather than forcing project teams into a single technology ecosystem.

“Bentley software can ingest and export data across different file formats, while connected environments such as ProjectWise allow information to move through the project without repeatedly returning to the beginning of the workflow,” Kumar says.

The next step is connecting those environments to AI.

Bentley has introduced Model Context Protocol servers for a number of its products, allowing software to connect at server level with AI models. This can support agentic workflows where different systems and AI agents interact to produce an outcome.

Kumar expects these workflows to generate significant time savings as the technology matures.

Building trust into AI

For the infrastructure sector to realise those gains, however, technology adoption needs to keep pace.

Trust is one of the biggest barriers. Engineering decisions can have significant safety, financial and operational consequences, meaning organisations need to understand how AI-generated outcomes have been produced.

Bentley’s approach includes what Kumar calls “transparency cards”, designed to make AI processes digitally traceable and auditable.

“If there are ever any issues with an outcome that was influenced by AI, you can transparently, and with trust, go back and audit it and see the digital traceability of that,” he says.

The engineer also remains responsible for the final decision, with checks and balances built into Bentley’s AI-enabled tools.

Data stewardship is another component of that trust. Kumar says Bentley’s end-user licence agreement states that customer data is not used to train its large language models unless the customer explicitly chooses to allow it.

“We don’t do it implicitly. Customers have to opt in to that,” he says.

For Kumar, those safeguards will become increasingly important as AI moves from assisting individual tasks towards managing more complex workflows.

Preparing the next workforce

Kumar sees solving the engineering workforce challenge as an ecosystem involving industry, government, universities and technology providers.

Bentley is working with universities including Griffith University, Western Sydney University and University of Technology Sydney to expose engineering students to digital engineering, model-first and asset-first approaches before they enter the workforce.

The aim is to reduce the change management burden that can accompany technology adoption later in an engineer’s career.

At the same time, Kumar cautions against viewing younger, digitally native workers as a replacement for the experience being lost as older engineers retire.

“You need this nice balance,” he says. “You’ve got people retiring, you’ve got new young people coming in and you’ve also got people coming in from overseas. This is a wide breadth of experience.”

That combination of experience and new thinking could provide an opportunity to reshape how engineering work is performed.

The foundation for that future is less glamorous than AI itself.

“The very first thing required to enable a technology-enhanced workforce, is clean, structured, connected data,” Kumar says.

He says poorly organised information can undermine even the most sophisticated AI system, recalling organisations where the same item has different names depending on who created the file, making it difficult for people and algorithms to find and interpret information.

Once information is structured and connected, AI can move beyond simply finding documents towards supporting agentic workflows, design outputs, operational readiness and decision-making.

The objective is ultimately to give engineers the information they need to make decisions, rather than requiring them to spend hours searching for it or repeating work that has already been completed.

To find out more, visit bentley.com



Source link