When machines learn to see

Greater safety at construction sites.

Occupational safety plays a pivotal role wherever heavy machinery is in operation and many metric tons of building material are transported. Driver-assistance systems, which have proven their roadworthiness in recent decades and have reduced the number of accidents, are increasingly popular on construction sites.

Uneven and sometimes non-drivable surfaces, complex obstacles, and difficult environmental conditions with strong vibrations and dirt pose new challenges for system developers.

To solve these challenges, ITK developed a system that detects objects and driving surfaces. This system embodies our core specialties of embedded, image processing, functional safety, integration, and verification & validation. The system meets the following requirements:

  • Recognition of people and vehicles
  • Identification of uneven driving surfaces
  • Determination of drivability based on incline information via modeling of the driving surface.
  • Warning of obstacles or people


The integration of the system depends on the degree of intervention in driving behavior. If the assistance system serves as a warning system that displays the surroundings with obstacles marked on a monitor – and uses acoustic signals to alert the driver to an imminent collision – then merely sensors, an arithmetic logic unit (ALU), and monitor need to be installed in older vehicles. On the other hand, if danger necessitates intervention in driving behavior – such as a brake booster or autonomous evasive maneuvers – then this affects the entire vehicle architecture, functional safety, and overall vehicle safety.

As a system supplier and integrator, ITK handles all these tasks to help improve occupational safety on construction sites.

To the business area