Motion analysis for optimizing processes

The working group “Scene Analysis” focuses on the optical recognition, analysis and interpretation of dynamic objects and persons as they interact with their environment. The human being is always at the center of the application. Besides human motion in general, specific movements of certain parts of the body, such as arms and hands, are also studied. The necessary data is acquired via imaging sensors, which are selected depending on the application and on the relevance of the various information for the overall scene. The sensors are deployed according to their technical capabilities (e.g. coverage, resolution, angle of aperture). The collected data is then placed in a higher-level context and processed, for example, for human-centered optimization of production processes. Quality objectives, efficiency improvements, ergonomics and employee training are possible applications.

Key areas of interest

  • Implementation of data-driven User Needs Assessments D²UNA
  • Activity and pose recognition for hand movements and people requires revision and inclusion of sensor data
  • Assistance system for transferring expertise: DIALOG reference project
  • Detection of falls and emergencies


The uses and applications are wide-ranging. The recognition of hand and arm movements - also in conjunction with the analysis of interaction with objects - is particularly suitable for assembly assistance, touch-free control or the intelligent classification of handling processes and interaction with machines. For example, person recognition is used to detect falls and emergencies, aid diagnoses or recognize emotions on the basis of typical movements of the human body.


Data-driven User Needs Assessments

Human-centered quality in the workplace can be achieved with the help of data-driven user needs assessments. Machine data and user data are linked as a basis for sustainable workplace design and system optimization.


Classifying hand and arm movements

Classifying hand and arm movements: Fraunhofer IPA develops technologies to automatically recognize typical everyday hand and arm movements.


Motion analysis for quality assurance

Thanks to optical motion analyses, complete production environments are analyzed, including all stationary and moving objects and people within it.


Safety for humans: Recognizing falls and emergencies

Scientists at Fraunhofer IPA have developed a safety system to automatically recognize emergency situations involving people and initiate the appropriate measures.


MonSiKo: An adaptive assistance system simplifies assembly in Industrie 4.0 processes

The project “Adaptive assembly assistance and interaction system using 3-D scene analysis and intuitive human-machine communication”, abbreviated from its German name to “MonSiKo”, focused on aiding assembly processes using modern sensor and communication technologies.



In the project “Self-learning and self-explanatory machine,” an intelligent and adaptive assistance system for operating and maintaining machines is being developed.



The project “Assistance system for manual workshop stations for people with disabilities using optical 3D sensors”, abbreviated from its German name to AMBOS-3D, was concerned with providing technical support to people with disabilities in workshops with the help of modern sensor systems and open-source developments.


PflegeCoDe: Care coaching to provide people with dementia with optimal support

Based on current technologies, an interactive electronic coach is being developed that combines technical, medical and organizational solutions.



The DIALOG project focused on helping new employees to learn new processes. An assistance system was used to aid and motivate them during the training process.