Technology

AI-based software for construction planning

21 October 2018 | By BIM+ staff

Software that uses artificial intelligence (AI) to help plan the best way to construct a building was one of the eye-catching start-ups to exhibit at Digital Construction Week in Excel, east London last week.

Developed by scientists from Stanford University in California, ALICE uses BIM models to automatically provide an array of options to construct the building based on parameters of scope and time, and cost.

In effect it augments the work of a contractor’s planner, says Brandon Young, Alice Technologies head of marketing.

“Working directly with your 3D BIM and using a simple rule-based system, ALICE explores millions of different schedule sequences and gives teams the power to answer “what if?” scenarios and completely reschedule in just minutes,” says Young.

“ALICE leverages human construction knowledge to power its AI engine and trials have shown it is able to reduce costs by about 15%.”

Alice Technologies started up about 12 months and comprises a team of experienced construction managers, software engineers, and computational methods specialists.

The software automatically produces detailed schedules and resource allocations, which accurately estimate the cost and time frame for a project. Users understand the construction process through an interactive visualisation, and can quickly tweak or modify a schedule to reflect the shifting realities of the construction cycle.

Alice Technologies has been working with companies in the US and is in talks with UK companies.

“Combined with our building expertise, ALICE provides the capability to evaluate millions of scheduling options that would take humans exponentially longer to accomplish,” says US-based Ricardo Khan, senior director of innovation at Mortenson Construction.

Design and engineering functions are increasingly being disrupted by AI as generative design is used to automate time consuming tasks, but construction and cost functions have been slower in coming to market.

Read related article