Human Detection System for Heavy Construction Equipment

Post by: Tobias Larsson
11th January 2026
No Comment

Safety increment without interference on productivity

Programme: Course project (add course code below in comment)

Course: MT2573

Corporate partner: VolvoCE

Completed: January 2026

Challenge: The challenge involved identifying hazardous activities on construction sites and developing a solution that maintains or increase the productivity of the worksite.

Solution: Unauthorized persons trespassing on construction worksites in crowded areas is a significant problem. This occurs when people take shortcuts instead of walking around the construction site. Combined with the limited visibility from inside excavator cabins and the many obstacles on the site, this can put construction workers out of the operators’ view, highlighting the need for a warning system.
The proposed solution is to install a system of cameras around the construction site, with screens inside the excavators that display live video feeds from these cameras. The video stream will be processed using image recognition software to identify workers on site and detect any potential trespassers. The system will alert the excavator operator with lights and sounds if a worker approaches the excavator or if someone enters the worksite without authorization.
This solution will not only enhance safety on construction sites but also help excavator operators stay informed about their surroundings. Additionally, it is expected to increase productivity by 30%. Currently, excavator operators often halt their work when they are unsure if the digging area is clear, requiring another worker to check. With the new system, operators can monitor their in-cabin screens, significantly reducing work stoppages and enhancing overall efficiency.

Impact: The increase in safety results in more comfortable working conditions and fewer fatal accidents in the construction sector. Additionally, the boost in productivity shortens the time required for construction projects, leading to faster development of new structures.

Prototypes: Using the Yolo model, we developed a software solution that differentiates between authorized site workers and unauthorized individuals on the construction site. The purpose of this project was to demonstrate the concept and showcase to excavator operators how such a solution could function. From this experience, we learned that excavator operators are very positive about the increased digitization of excavators and the introduction of more technology in their equipment.

Project team: