Search Results

Monitoring driver’s postures has extensive applications. The postural information could be used for the development of smart airbags, for detecting possible fatigue in long travel and for recognizing activities which may determine if the driver has enough time to take over the control in an intelligent vehicle when encountering hazardous situations. Microsoft Kinect is one of the best candidates for monitoring driver’s posture thanks to its innovative feature of real time motion capture without use of markers and its low cost. However, when body parts are partially occluded, the accuracy of Kinect data will drop markedly. Inspired by previous researches, the present work focused on testing a data driven approach for improving driver’s upper body movement reconstruction with a Kinect camera. Firstly, we organize a database of accurately captured driver poses from different motion clips with filtered structures, including a global motion graph and multiple local motion graphs.

An airplane passenger seat, like other seats in transportation, is used by thousands or millions of people. The seat should be designed to accommodate the maximum number of a target population by taking into account the variability of body size and also the environment’s constraints. Thanks to a fully adjustable experimental seat recently built at IFSTTAR, data of the preferred seat profile and compressed seat pan surface were collected in function of seat pan and backrest angle from a sample of 36 differently sized participants. Parametric models were obtained to predict optimal seat profile parameters in function of a sitter’s anthropometric characteristics, seat pan angle and seat back angle. Using a population simulation approach, a sample of 500 males and 500 females were generated randomly based on the distribution of relevant anthropometric dimensions.

A new experimental seat was designed to investigate sitting biomechanics. Previous literature suggested links between sitting discomfort and shear force, however, research on this topic is limited. The evaluation of sitting discomfort derived from past research has been primarily associated with seat pressure distribution. The key innovative feature of the experimental seat is not only pressure distribution evaluation but shear forces as well. The seat pan of the experimental seat compromises of a matrix of 52 cylinders, each equipped with a tri-axial force sensor, enabling us to measure both normal and tangential forces. The position of each cylinder is also adjustable permitting a uniform pressure distribution underneath the soft tissue of the buttocks and thighs. Backrest, armrests, seat pan and flooring are highly adjustable and equipped with forces sensors to measure contact forces.

This paper presents a pilot study aimed to carry out a method for prehension movement recording. This method consists of using, simultaneously, a motion analysis system (Vicon®) and a sensitive glove (Cyber Glove®). The objective is to be able to give a set of complete data to our hand model in order to simulate the human prehension movement. In a first study, only a data glove was used to record grasping movement. Many problems appeared when analysing the output of this system: 1. the hand model considered in the glove software is different from our model, 2. the arch degree of freedom does not exist in the glove model, despite the existence of the corresponding sensor, 3. the output from ab duction and flexion sensors are linked, 4. the calibration of the thumb is not efficient. From these technical constraints, another movement recording system - motion analysis system - is used and data glove completes the data recorded with motion analysis system.

In addition to three main commercial human models (Ramsis, Jack, Safework), there exists also many non-commercial ones used in research laboratories. Sometimes, it becomes necessary to convert motion data from one model to another. This occurs in the case of the European research project REALMAN. In this project, three models (Ramsis, Jack and Man3D) are used by different partners. As they have different kinematic architectures, we have to write converters allowing collected motion data to be used by all three human models. In this paper, we present the method developed for converting the data from the Human Solution's Ramsis to Man3D, a noncommercial model developed at INRETS. We have shown that it is possible to convert motion data from one manikin to another using a method based on skeleton model comparison.

Numerical models are used more and more to visualize a human operator within a work environment and simulate his movements. Many models are limited in their ability to simulate complex activities like prehension and objects manipulation. The hand models proposed in the literature are relatively simple, especially assuming the palm as a rigid body, which leads to unrealistic representations of complex attitudes. The objective of the present study is to develop a more advanced hand model, able to properly simulate prehension postures. A 25 degrees of freedom (DOF) hand model has been proposed including 2 DOF for representing the palm arch. Compared to the model without palm arch, the proposed model has made significant improvement of the hand posture representation, suggesting the need of including palm arch for simulating complex hand grasping attitudes.