Person detection in human-machine-interaction

Project title

Opportunities and risks of Ambient Intelligence (AmI) from a safety point of view. Sub-project 2: Reliable Person Detection in Human-Machine Interaction

Funding

Federal Institute for Occupational Safety and Health (BAuA)

Motivation

The term of “Ambient Intelligence (AmI)” summarizes activities in research and development that extend life and work environments with intelligent functions. A main element of those techniques, that are associated with “Industry 4.0”, is the connection between autonomous, self-controlled, self-configured, knowledge-based, sensor-aided and spatially distributed production resources (machinery and equipment), including systems for planning and control. All these systems include sensor configurations, whose measurements are processed by computer processors, causing directed actions. The application of Ambient Intelligence and Industry 4.0 systems is the topic of several research projects. Systems for human detection are AmI-based security solutions, that allow a close cooperation between human beings and machines in production processes without separating safety fences. This leads to a freedom of movement for the machine operator and to a high flexibility in work processes. Person detection systems are therefore essential applications of AmI-based safety systems and part of Industry 4.0 systems. To guarantee a safe interaction between humans and connected, sensor-aided and computer-operated machines inside the work process, it has to be assured that a human detection system detects the human body reliably. The result of the human detection has to be transformed into control parameters for the planning and controlling of the machine’s motion sequences. The specific characteristics of different measurement principles and the environmental properties lead to possibilities and limitations for the application of these measurement principles. It is essential for a possible user to show him these limitations and possibilities to guarantee a safe operation of human detection systems.

Aims

The aim of this sub-project is to classify fields of application and limits for reliable person detection in human-machine interaction. Based on research in literature and on the market, a well-structured application matrix will be elaborated. The matrix evaluates different measurement principles for human detection, concerning fields of application, technical parameters and environmental factors. It therefore supports the work of company’s planners and purchasing managers.