Robotics researchers at the University of California in San Diego (UC San Diego) are using new mapping and visual technologies to position robotics to take on jobs in construction that are “dirty, dull and dangerous.”

UC San Diego Electrical and Computer Engineering professors Truong Nguyen and Nikolay Atanasov and their team of researchers at the Jacobs School of Engineering are “exploring ways to integrate robotics into the construction industry and improve 3D construction and mapping,” according to a release issued by the department. The grant-funded project runs through August 2027 and is supported by the Korea Institute for Advanced Technology, the Korea Electronics Technology Institute and IT construction solutions company ITOne, India.

Harsh settings or delicate duties

Atanasov, who directs the university’s Existential Robotics Laboratory, said the project’s goal is to automate dangerous and dirty aspects of the industry that could be difficult or perilous for human construction workers. These tasks may include lifting heavy objects, welding or working in areas that may be unsafe for humans.

The global market value for industrial robot installations reached an all-time high of $16.5 billion, according to the International Federation of Robotics, which recently identified the top robotics trends of 2025. One significant trend from the IFR report is robots addressing the labor shortage in manufacturing, with these devices undertaking tedious jobs such as visual quality inspection, hazardous painting or heavy lifting.

While seeing robots in manufacturing, industrial and even customer-facing service settings is becoming more commonplace, UC San Diego researchers are taking steps with technology to help robots understand and adapt to the construction industry environment and the specific actions required to fulfill tasks.

“To change a light fixture on a high warehouse ceiling, for instance, the robot must synthesize information on its surroundings with potential actions that are safe to perform to achieve certain goals. And that assessment must somehow be accessible to the human watching or guiding it,” the researchers reported.

To continue to acclimate robots to the construction environment, the researchers are using models and mapping to assess their environment.

“Mapping essentially turns an object the robot sees into a 3D reconstruction that can be viewed and manipulated on a computer screen,” said Runfa Li, an electrical and computer engineering PhD student in Nguyen’s lab and the head of mapping efforts. Although this can be accomplished using various techniques, their method involves specific camera poses and orientations. “If we have images of an object and know which direction the camera is facing, we can accurately reconstruct the surface mesh.”

The next step involves human interaction and high-level instruction that directs the robot, such as clicking on the part or object the robot must leverage. Then, the robot translates and executes commands to complete the task.

“It turns into a planning and control problem where the robot needs to take into account its understanding of the 3D environment, its safety and the human’s wants, and then synthesize that into a plan,” Atanasov said. “It’s a major task; we’re interested in further developing synthesized robotics that streamline the process and make human-robot interactions smoother.”

Visualization technology

With visual mapping and 3D reconstruction technology, robots can ascertain safe movement or engagement with other objects, and the ability to move safely in an unstructured environment is exciting and opens new possible applications, according to Brian Lee, who is part of Atanasov’s lab. “The robotics systems we develop are able to adapt and evolve based on the people and situation—the context—they encounter,” Lee said. Atanasov’s lab is working alongside the UC San Diego Contextual Robotics Institute

In construction, the flexibility and awareness from visualization can be used for building projects with modular design and prefabricated components, for example. Mapping and processing technology will also lead to more high-quality virtual reality using detailed scanning and camera snapshots to instantly and accurately reconstruct a scene.

The technology will be tested in a multifamily construction environment using robotic arms, but Atanasov said use-cases can expand to include hospitals, warehouses, autonomous vehicles and agriculture.