Fraunhofer Institute for Factory Operation
and Automation IFF
Fraunhofer Institute for Factory Operation
and Automation IFF
© Fraunhofer IFF

Measurement and Testing Directly in the Production Process

© Fraunhofer IFF

Research and Development in Measurement and Testing Technology

Visual Assistance

Assistance and Quality Inspection for Manual Manufacturing Processes

Typically, processes and operations of advanced systems for mass production are highly automated.
Cognitive problem-solving skills, flexibility and adaptability make human labor superior to automated systems when small lots or prototypes are being produced. Assistive systems support workers in complex and widely varying operations by providing instructions and assistance for a given situation. Their combination with optical quality inspection systems produces integrated process chains with guaranteed quality and reliability. The elimination of defects and the extra work they generate boost value added.

Criteria for the Use of Visual Assistive Systems

  • Manual labor is performed.
  • Operations are complex and variable.
  • Product quality documentation is required.
  • Components with relevance to safety are manufactured.

Typical Fields of Application

  • Manual assembly
  • Servicing and maintenance
  • Order picking
  • Packing
  • Orientation and alignment

Visual Assistance

The approach to generating an assistive function is based on an augmented reality system. Manual operations are supported by instructions and information provided by a visualized real work scene that is augmented or overlaid by additional graphics and texts.

The work area is monitored by a video camera and, in the simplest case, visualized on a monitor in a worker’s field of vision. The camera can be stationary or record a scene from variable perspectives. CAD models, e.g. of a unit assembled manually at a workplace, are behind the system. Referencing methods are used to determine the camera’s perspective on the work scene from the camera images. This orientation information is used to align the CAD model, thus generating an identical view of the scene synthetically.

The real camera image can be overlaid with individual components of the synthetically generated scene to generate instructions and thus provide intuitive support and assistance.

The real camera image can be overlaid with individual components of the synthetically generated scene to generate instructions and thus provide intuitive support and assistance.

For instance, components, including their correct location and alignment, as well as their numbers can be presented in the sequence they are assembled manually. The sequence of set or adaptive steps is stored in the system beforehand.

Quality Inspection

Camera images used to evaluate the progress of work can also easily be used to inspect the objective quality of manual work.

Classic image processing methods can be used to evaluate results by comparing them with CAD data. The setup can be expanded, e.g. with more cameras or structured lighting, to generate 3D data of the scene and thus another option for inspecting quality.

Research and Development in Measurement and Testing Technology