International Journal of Human Factors Modelling and Simulation (5 papers in press)
Pedestrian Simulators for Traffic Research: State of the Art and Future of a Motion Lab
by Ilja Feldstein, Christian Lehsing, Andre Dietrich, Klaus Bengler
Abstract: While classical driving simulators have for decades been a valuable tool for the investigation of human behavior and validation of advanced driver assistance systems (ADAS), the development of pedestrian simulators is still in its beginning stages. However, with an increasing complexity of ADAS, it is not sufficient anymore to evaluate those uniquely from the perspective of a driver using driving simulators. Mainly based on motion tracking and virtual reality (VR) technology, pedestrian simulators are used to investigate human behavior from the pedestrian perspective especially in urban traffic scenarios in a reproducible, safe, and cost-efficient way. They enable research on pedestrian behavior in potentially hazardous traffic situations (e.g., crossing scenarios). Becoming an increasingly valuable tool for car manufacturers and original equipment manufacturers (OEMs) in the development process of ADAS, e.g., pedestrian detection and avoidance, the importance of these simulators will grow given the emergence of autonomous and silent cars (e.g., electric vehicles) in the near future.
This paper will help researchers starting off in this research field to get a general idea and also some primary technological input. It explains the potential and possible areas of application of this particular simulator paradigm and gives a detailed overview of some of the technologically most advanced virtual reality simulators used by various research institutes for pedestrian investigations. More specifically, the technology used in the pedestrian simulator developed at the Technical University of Munich is thoroughly described, and components such as head-mounted display, motion capture system, and simulation software are discussed.
Finally, the approach of linking a driving simulator to a pedestrian simulator so that both participants can meet simultaneously in the same virtual environment is introduced. This promising approach facilitates social interaction regarding mutual behavior adaption in the virtual environment of a driving simulation. Experimental evaluations that use this multiple simulator setup can address a broader variety of research questions in traffic-related areas where the interaction between different classes of road users can be assessed as opposed to the reaction to programmed agents used in conventional driving simulators. Regarding data validity, especially in urban scenarios where interaction plays a significant role, the approach of two human beings encountering each other in a safe and reproducible traffic environment shows its potential.
Keywords: pedestrian simulator; virtual reality (VR); head-mounted display (HMD); motion capture; road crossing investigation; linked simulation; social interaction in traffic.
The Effect of Task Characteristics on the Choice to Lean, Upper Body Postures and Joint Loading During Simulated Automotive Manufacturing Tasks with One-Handed, Submaximal Exertions
by Kayla M. Fewster, James R. Potvin
Abstract: The purpose of this study was to investigate the preferred leaning postures, while giving participants the choice to lean when completing tasks with constrained reaches. Twenty female participants completed a variety of different exertions with and without a leaning surface available. The frequency of choice to lean changed with task hand location. The long reaching task hand locations resulted in the most frequent choice to lean, and this decreased trunk and task arm shoulder loading, while allowing the participant to get closer to the task. These findings will be of use to industry in validating future leaning posture prediction software to help guide leaning posture estimates during proactive risk assessments.
Keywords: Leaning; Proactive Ergonomics; Constrained Posture; Automotive Manufacturing.
Examination of robotic manipulability indices to evaluate upper limb manipulability in digital human models
by Hiroshi Endo
Abstract: Robotic manipulability indices (RMIs) quantitatively evaluate kinematic factors regarding manipulability and can be applied to manipulability evaluations in digital human models (DHMs). However, in the kinematics of human motion based on Eulers angles, the rotational transformation order around three orthogonal coordinate axes is arbitrary, meaning that the coordinate transformation at a shoulder joint is not unique. Because RMIs are calculated from the coordinate transformation matrix, the RMIs of an upper limb may be influenced by this arbitrariness. This study examined the effects of the rotation order of the coordinate transformation at a shoulder joint on the manipulability evaluation of an upper limb. The results indicate that RMIs depend on rotation order. Thus, a modified method that eliminates the influence of rotation order was subsequently proposed. Experimental results indicate that RMIs calculated via this proposed modified method are rotation order independent, verifying that it contributes to manipulability evaluation in DHMs.
Keywords: DHM; Digital human models; Manipulability evaluation; Manipulability ellipsoid; Workspace; Reaching; Upper limb movements; Shoulder joint; Kinematics; Rotation transformation of coordinate system; Euler’s angle; Ergonomics.
Ergonomic Assessment of a Physical Task Using Two Different Digital Human Modelling Systems: A case study
by Marisol Quintero-Duran, Gunther Paul
Abstract: The cost of occupational musculoskeletal disorders (MSDs) causes a significant burden on the health system, and lower back pain (LBP) is associated with a substantial portion of MSDs. Australia has a high prevalence of MSDs for health care workers, such as nurses. Two digital human models (DHMs) Siemens JACK and imk EMA were used to investigate if hospital bed pushing, a simple task and hazard that is commonly associated with LBP, can be validly simulated and ergonomically assessed in a virtual environment. While JACK and EMA have implemented a range of common physical work assessment methods, the simulation of a dynamic task such as bed pushing remains a challenge. This research highlights limitations in the DHMs studied, and the need for further research in this area. In particular, the implementation of legacy two dimensional, low resolution ergonomic methods in an analytic, high resolution three dimensional software system is critically reflected.
Keywords: Siemens JACK; imk EMA; workload assessment; hospital bed; health care; nursing; pushing task; musculoskeletal disorder; physical workload; Digital Human Model; virtual environment.
Simulation Based Discomfort Assessment of Two-Wheeler Riders
by Mohd Parvez, Abid Khan
Abstract: Ride discomfort apart from various other factors is caused due to transmission of vibrations to human body while driving. Road surface irregularities or road profile act as a major source of vibration generation to which a vehicle rider is exposed. In the present study, a 6DOF simulation model of two-wheeler coupled with rider was developed considering road profile as a major source of vibration generation. The model developed may be used to assess ride discomfort of rider caused due to WBV exposure for various type of roads classified as per ISO 8608:2016. The model was developed using LabVIEW graphical software Control Design and Simulation module. The model developed was validated for a two-wheeler named HERO SPLENDOR PRO on all the three types of road profile i.e. good, average and poor at a vehicle speed of 20km/h (since the distance of particular road was very short hence more data could be acquired for better validation at lower speed ). The weight of the vehicle rider was 85kg. Assessment of discomfort, according to ISO 2631-1:1997, was done for 85kg rider on Good road profile at 40km/h vehicle speed by calculating RMS acceleration at seat and it was found that the ride was uncomfortable (0.977756m/s2 RMS acceleration). Further to investigate the effect of road profile, vehicle speed, seat stiffness and seat damping on RMS acceleration (level of comfort) at seat, a full factorial experiment with three levels for each (34) was performed. It was found by performing ANOVA that there was significant effect of all these factors as well as their interactions on RMS acceleration. Also from full factorial results, minimum level of vibration (0.506m/s2 RMS acceleration) was found for Good road profile, 20km/h vehicle speed, 10000N/m seat stiffness and 300N-s/m seat stiffness which predicted that the comfortable ride can be achieved by driving two-wheeler on Good road profile at low speed with a seat of low stiffness and high damping values.
Keywords: Two-wheeler vibration; Random road profile; Ride comfort; Whole Body Vibration.