Buildots boosts digital twin process mining with $30M

Buildots, a construction digital twin company, garnered a $30 million series B round led by Lightspeed Ventures, bringing its total investment to $46 million. Buildots will use the new funds to double the size of its global team, focusing on sales and R&D to expand its digital twins efforts, which use process mining techniques to improve outcomes as construction trades go digital.

“The new funding will support our ambitious growth plans for 2021-2022, including extending our existing sales team and opening new territories,” Buildots cofounder and CEO Roy Danon told VentureBeat.

“It will also support additional enhancements to the product, such as supporting more project workflows, integrations with other ecosystem players, and [fine-tuning] our AI to provide more critical insights to our clients,” he continued.

Buildots has early customers in 13 different countries, including Build Group in California and Washington state, MBN in Germany, Gammon in Hong Kong, and Wates in the U.K. Previous investors include TLV Partners, Future Energy Ventures, and Tidhar Construction Group.

Operationalizing digital twins

While other companies focus on the design or presentation of 3D construction data, Buildots specializes in operationalizing it. Buildots has concentrated on the gap between existing tools for design, scheduling, document management, and process controls that provide visibility into what’s happening on construction sites. The company focuses on higher-frequency updates and greater detail.

Founded in 2018, Buildots aims to improve the user experience for workers and managers. Its special sauce lies in streamlining and automating the reality capture process using hardhat-mounted 360-degree cameras.

The Buildots tools bring process mining techniques to construction projects. The software is able to track the exact process by which construction projects are built for the first time, Danon said. Connecting these process models with the original design and schedule information is intended to help managers learn more about bottlenecks in their existing processes and how to get them right the first time.

In the background, Buildots’ AI algorithm double-checks new work against the plan, tracks progress, and updates an as-built digital twin model. The granularity of information in Buildots enables teams to drill down on any issue found on-site and take immediate actions to keep the project on budget and on schedule.

Identifying bottlenecks through process mining

A project’s current state is captured on an ongoing basis through cameras while teams make incremental changes. Proprietary AI and computer vision algorithms fuse this data with the latest design and scheduling plans and update the platform’s internal digital twin.

For example, one European company using Buildots discovered that its concrete finishing team was proceeding much more slowly than the partition building team. This created a bottleneck for the construction of new floors.

The Buildots application alerted managers to the problem. Then it helped them formulate a new plan that diverted workers away from building partitions to finishing the concrete, which reduced delays for everyone.

Improving 3D model quality

The platform can also identify quality gaps between the plan and what was actually built. It is common for humans to miss some elements when manually comparing building documents to what they see.

Manual tracking processes tend to be infrequent; have low granularity; and rely on people’s objectivity, skill, and attention to detail. Once such processes are automated, teams capture details more frequently, which reduces the delays in resolving problems. It is also possible to drill down into construction progress at the level of an individual socket and its different stages of installation.

For example, the two images below show a 3D model of the plan on the right and a white outline where the application detected a missing outlet.

“While this isn’t a huge deal for any given outlet, on the average project, we spot a missing element for every 50 to 100 square meters,” Danon said. Averting hundreds of those issues can lead to a substantial efficiency improvement.

Above: Here, software detected an overlooked power outlet requirement.

Image Credit: Buildots

Transparent AI builds trust

The focus on updating and auditing the data trail across the lifecycle of a project is another key feature. Existing market solutions such as PlanGrid, Procore, and others have already paved the way for construction teams now using mobile apps on the construction site. Today’s engineers and managers are generally comfortable using iPads or web applications in their day-to-day work.

But all these tools require someone to enter data manually. In contrast, Buildots’ approach to digital twins automates this process and connects the data to an audit trail woven into AI models. This transparency allows construction teams to understand how conclusions about a particular project scheduling problem were reached.

“We have built our platform with the principle of transparent AI, meaning that every conclusion the system makes can be drilled down into so that construction managers can develop trust with their new virtual team members,” Danon said.