Fraunhofer Institute for Industrial Mathematics ITWM
Fraunhofer Institute for Industrial Mathematics ITWM
Montage Automotive
© iStockphoto
Die von uns entwickelten Methoden und Algorithmen dienen dazu, dynamische Bewegungen von Arbeitenden in der Pro- duktion besser als bisher ergonomisch zu bewerten. Wir schauen uns speziell Mon- tagetätigkeiten in der Automobilindustrie an.

The methods and algorithms developed by us are used to asses dynamic motions of workers in production w.r.t. ergonomic aspects better than before. We are especially looking at assembly activities in automotive industry, like the assembly of an engine shown in the picture.

© iStockphoto

Digital Human Models and Human-Machine Interaction

Solutions for Driving Simulators

Numerous high-end products continue to be manufactured by people (and will continue to be so in the near future). Wherever dexterity, intelligence and experience are indispensable for high product quality, people are on duty.

The digitalization of the product development process requires simulation tools. These should map human work activity sufficiently well that the influence of the »human factor« on product quality can be mapped and product quality ensured. The ergonomic design of workplaces is equally important, supports experienced workers in maintaining their health and in this way likewise supports a high product quality.

Cost-effective due to virtual modeling: Simulation tool enhances ergonomy of industrial workplaces.
© Industrial Path Solutions Sweden AB

Human Models for the Digital Factory

With the digital manikin IPS IMMA of our divison »Mathematics for Vehicle Engineering« can be used to simulate human-machine interactions during assembly work. Our software module IMMA offers fast and efficient algorithms for easy evaluation of the ergonomics of assembly processes. The simulated movements of an IMMA manikin result from a biomechanical model. They are computed in such a way that collisions with the objects in the environment are avoided and the motion of the whole manikin is performed as comfortable as possible. To achieve this we optimize »biomechanical comfort functions« which allow us to reliably predict human movements even for completely unknown and new scenarios.

 

Multi-Body Model Calculates Optimal Movements

As a biomechanical model, we use a multi-body model derived from a simplified human skeletal structure. The model consists of 82 bone segments connected by joints with a total of 162 kinematic degrees of freedom. The user instructs the manikins with an instruction language to work in different postures and interact with the environment. Based on the biomechanical multi-body model, human-machine interactions are automatically calculated in such a way that they not only avoid collisions, but also account for the following influences:

  • equilibrium of the forces and moments acting in the model,
  • kinematic constraints,
  • external contact forces,
  • posture comfort.

We then further analyze the results of such a motion simulation w.r.t. ergonomic criteria. In IMMA several manikins with different anthropometric datas can be computed simultaneously. This makes it easy to perform simulations that also take into account the variance of human body characteristics within a population.

Action Models and Human-Machine-Interaction

The human being plays a decisive role not only as an actor in a working environment, but also as an operator of a vehicle or a mobile machine. How he solves the work task has a significant influence on the physical strain that the person experiences. Operator or action models are mathematical descriptions of actions including decision-making processes that enable the simulation of such action sequences and their optimization according to ergonomic criteria.

Our research work on »action models« aims at the modeling and simulation of human actions influenced by perceptions of the environment, as well as coupling the typically time-discrete action models with models of system dynamics. 

A coupling of action and digital human models opens up the possibility:

  • to implement roughly specified tasks in action sequences (e.g. »take a certain component, assemble it at a certain location« as part of an assembly process),
  • to calculate individual partial steps of such a sequence by means of optimum control,
  • to categorize actions according to given criteria, e.g. as economically efficient, ergonomically favorable or as health-friendly as possible.

With this human-machine interaction driving simulator we can systematically analyze many work processes and optimize them, e.g. with regard to physical strain.