Matt Reed, Ergonomics and Safety Research

Matthew P. Reed, Ph.D.

I'm the Don B. Chaffin Collegiate Research Professor and Head of the Biosciences Group of the University of Michigan Transportation Research Institute. I conduct research in a variety of areas relating to anthropometry and biomechanics, including vehicle ergonomics and vehicle occupant crash protection. I'm also a Research Professor in the Center for Ergonomics in Industrial and Operations Engineering, where I lead the Human Motion Simulation Laboratory. The HUMOSIM Lab develops movement simulation algorithms and ergonomics analysis tools for use with digital human modeling software. I am also a Research Professor in the Integrative Systems+Design program in the College of Engineering, where I serve as Acting Director of the Design Science program.

Follow the links at the right for more information about my research and see these highlights of recent projects.

Google Scholar Profile

Scopus Public Profile (ID: 7401800235)

ORCID ID: 0000-0001-6650-0727

University of Michigan Researchers

Other affiliations:

Associate Editor, Traffic Injury Prevention

Associate Editor, SAE International Journal of Transportation Safety

Associate Editor, International Journal of the Digital Human

International Research Council on the Biomechanics of Injury

Update: 2017-12 JapaneseComparison

Designing vehicles to be used in multiple markets is challenging in part due to differences in the body sizes across nations. Yet, little research has been published on national differences in body size, shape, and posture for vehicle occupants. We conducted a focused study comparing driving posture and body shape for Japanese and U.S. nationals. The U.S. data was drawn from a large-scale study of driver and passenger posture. We recruited Japanese nationals who had been resident in the US for less than one year and measured them in the same conditions. The results show small but significant differences in driving posture after accounting for body size. Body shape differences were more apparent. These results are valuable for companies designing vehicles for both markets.

Update: 2017-11 Scarab

My colleague Dr. Jingwen Hu presented a paper at the Stapp Car Crash Conference on our extensive collaboration with ZF on rear-seat occupant protection. This research, funded by NHTSA, aimed to identify the extent to which rear-seat occupant protection could be improved by adding previously developed technologies. Historically, safety technologies used routinely in front seats, such as airbags and advanced seatbelts with pretensioners and load limiters, have been installed much less frequently in rear seats. Consequently, improvements in occupant protection in these seating positions have been slower than for the front seat. This study documented substantial improvements that could be made, particularly through the use of novel airbag designs.

Update: 2017-10 Passenger Lap Belt Fit

I attended the annual meeting of the Association for the Advancement of Automotive Medicine in Las Vegas, NV this month. I presented a paper comparing the belt fit obtained by rear-seat passengers with the belt fit measured on Hybrid-III ATDs using standardized installation procedures. [This paper has been accepted for publication in a special issue of Traffic Injury Prevention.] On average, the belt fit for the ATDs was much lower and more advantageous, relative to the pelvis. The shoulder belt also routed much closer to the centerline for the ATDs. These differences in belt fit may have important implications for the design of restraint systems. The lower placement of the lap portion of the belt likely means that the ATDs are less likely to submarine than the ATDs. This in turn may result in manufacturers choosing flatter belt angles than would be optimal for human occupants. Similarly, the more-centered shoulder belt placement may result in an underestimate of the chest loading that would result in a forward crash, diminishing the importance of load limiting. Further research with parametric human body models is needed to evaluate these effects.

Update: 2017-10 MATT

I attended the annual conference of the Human Factors and Ergonomics Society in Austin, TX this month. Justin Haney, a PhD student working with advisor Clive D'Souza in the Human Motion Simulation Lab, presented recent work on measurement and modeling of threading operations. I participated in a panel discussion on the use of virtual fit testing (VFT) to improve the accuracy and flexibility of accommodation assessments of office workstation designs. My colleague Matt Parkinson and I have developed a new database of US civilian anthropometry by weighting the US CAESAR data, augmented with additional high-BMI data from UMTRI research. We've built a spreadsheet for VFT around the new weighted database, providing a simple interface to obtain accurate multivariate assessments of office workstation accommodation. HFES will be making this new VFT tool available along with a revised version of the ANSI/HFES 100 office workstation design standard.

Update: 2017-09 2YO Belt Fit Manikin

Our group was well represented at the annual European meeting of the International Research Council on the Biomechanics of Injury. We presented two full papers and one short communication describing recent work.

Reed, M.P. and Boyle, K.J. (2017). Development of a manikin representing a two-year-old child for belt-fit measurement. Proc. 2017 IRCOBI Conference. Antwerp, Belgium.

The seats in commercial aircraft in the US are required to provide appropriate lap belt fit for all passengers ages 24 months and older. However, no belt fit procedure or measurement device is specified. This paper describes an internally funded effort to develop a belt-fit test device representing a typical two-year-old child. The external shape is based on output from a statistical model of toddler body shape. The external surfaces were created using 3d printing and mounted to an internal armature that includes hip joints and an articulate four-bar neck. The next step is to validate the belt fit measures obtained using this manikin using measurements from children. As far as we are aware, this is the first belt fit device to be created using 3d printing, as well as the first to be created using a statistical body shape model.

Jones, M.L.H., Ebert, S.M., Hu, J., and Reed, M.P. (2017).  Effects of high levels of obesity on lap and shoulder belt paths. Proc. 2017 IRCOBI Conference. Antwerp, Belgium.

Belt Paths

Analyses of crash data indicate that high BMI is associated with a higher risk of thoracic injury in frontal crashes. In previous work, we found only small differences in shoulder belt placement relative to the clavicle associated with obesity, although large effects on lap belt placement were found. In the current paper, we used functional regression techniques to model the effects of driver attributes on the path of the belt relative to the skeleton. As expected, higher BMI is associated with with belt routing farther from the skeleton, but we also found sex differences for both the lap and shoulder belt after holding stature and BMI constant. In general, we find that higher BMI causes the lower torso portion of the belt to route more laterally, which may cause increased risk of rib fractures and liver injuries for drivers. We will use these results to improve the routing of belts in simulations of frontal crashes with high-BMI drivers.

Hu, J., Zhang, K., Fanta, A., Jones, M.L.H, Reed M.P., Neal, M., Wang, J-T., Lin, C-H., and Cao, L. (2017). Stature and body shape effects on driver injury risks in frontal crashes: a parametric human modelling study. Proc. 2017 IRCOBI Conference. Antwerp, Belgium.


My colleague Dr. Jingwen Hu presented results of ongoing work examining the effects of body size, shape, posture, and position on crash outcomes. The parametric modeling paradigm developed in our group allows us to morph high-fidelity finite-element human models to represent any of a wide range of occupant sizes and shapes. In this work, we used validated posture-prediction models to position six human body models: three representing the reference anthropometry of the small-female, midsize-male, and large-male ATDs that are widely used for vehicle safety system assessment, and three models with the same stature but a BMI of 40 kg/m^2 (roughly the 95th percentile for US adults). The results showed very different restraint system interactions across the body size and shape space. BMI was associated with higher injury risk, consistent with the field data, and both short and tall drivers were at higher risk. The results highlighted the possibility of the driver's torso disrupting the deployment of the airbag, reducing its effectiveness. Future studies with a much larger number of body sizes and shapes are underway.

Update: 2017-08 WIAMan

The US Army has taken delivery of the first batch of "Gen 1" Warrior Injury Assessment Manikin (WIAMan) ATDs. This is a major milestone in this important program to create an entirely new physical test device for vehicles exposed to underbody blast. Our group developed the anthropometric specifications for the ATD (size, shape, posture) and has been active with other universities in developing the performance specifications and injury criteria through our drop-tower test program. More info on that research is available elsewhere in these pages.

Update: 2017-08Braking Motion

We've gathered a large set of data on the responses of front-seat passengers to unexpected hard braking and other abrupt vehicle manuevers. These data will be valuable for developing and tuning active human body models that can simulate occupant responses during crash-avoidance actions triggered by vehicle automation. Pilot data we gathered last year showed that head excursions exceeding 200 mm were observed during hard braking, even with normally positioned passengers who are properly belted. Movements of this magnitude may have important implications for the design of restraint systems. The primary focus is determine whether restraint systems, including seat belts and airbags, need to be improved to handle occupants who are out of the expected position.

These videos were created using data from a single Microsoft Kinect sensor mounted in the vehicle cabine. We are using methods developed at UMTRI to fit subject-specific avatars to these data to obtain accurate head and torso kinematics during the events.

Lane Change Motion

Update: 2017-07 cervical spine model

The cervical spine is of considerable interest in biomechanics research due to its complexity and its importance as a site of both low-severity and high-severity injury (e.g., whiplash-associated disorders and fractures with spinal cord injury). My colleague Dr. Monica Jones and I recently published a technical report describing new two- and three-dimensional parametric models of cervical spine geometry and posture. The data for the 2d model were drawn from a Snyder et al. x-ray study conducted at UMTRI in the 1970s. Using the same methods we've previously applied to model body shape, we created a parametric model that generates sagittal spine geometry as a function of sex, stature, age, and head-to-thorax posture. This is the only parametric model of its kind for unsupported head postures typical of seated environments, such as vehicles. In addition, we linked this 2d model to 3d geometry provided by collaborators at Johns Hopkins University Applied Physics Lab (APL). Using 3D spine geometry data extracted by APL from medical imaging (CT) studies, we created a parametric model that predicts 3d bone geometry from 2d shape. One unexpected observation is that lateral dimensions of cervical vertebrae are essentially uncorrelated with sagittal dimensions. The resulting 3d model can be used as input for finite-element models used to examine the influence of posture and spine shape on injury risk in a wide range of exposures. This model is implemented in Python and is freely available for use -- just contact me to get it.

Update: 2017-07 error distribution

My colleague Dr. Daniel Park and I recently completed a research project for the U.S. Army that tackled an array of cutting-edge problems relating to human body measurement and modeling. The report on this work is now available online. From the abstract: "Two metrics were devised for comparing two body shapes represented by surface meshes. The distance from the nodes of one mesh to the polygonal surface of another was defined as mesh error. Six torso dimensions computed between mesh nodes that are analogous to standard anthropometric measures were compared to compute mesh error. Analyses were performed using three datasets: 236 male Soldiers, 200 Air Crew, and 73 civilian women. Statistical body shape models (SBSM) were developed using methods developed and adapted in previous UMTRI research. A standardized template was fit to each scan to enable the analysis. Mesh error was found to diminish smoothly with the number of PCs used for reconstruction, with minimal improvement after 100 PCs. When conducting regression predictions, retaining more than 80 PCs provided minimal improvement in mesh or dimension error metrics. A simulation study demonstrated that improvements in regression model performance when using more than 50 subjects were small. Errors in predicting Air Crew torso mesh dimensions using 10 standard anthropometric variables averaged less than 10 mm. A novel method was developed to predict seated body shape from standing body shape, and a new inscribed fitting method enabled generation of accurate avatars from scans of individuals wearing clothing and gear. A pilot test demonstrated the potential for scanning prone individuals using a transparent table to obtain good coverage." Whew! Lots of good stuff in there. Contact me if you'd like details or would like to use these models or methods.

Update: 2017-06 boundary manikins

My colleagues Dr. Monica Jones and Dr. Daniel Park represented our group at the International Digital Human Modeling Symposium in Bonn, Germany this month. Dr. Jones presented methods and preliminary results from our work on occupant dynamics during abrupt vehicle maneuvers. The goal of this work is to understand how occupant posture and position many change due to automatic crash avoidance systems so that restraint systems can be designed appropriately to protect occupants when a crash does occur. Dr. Park presented an investigation into the development of boundary manikins. Principal component methods are widely used for developing small families of 3d manikins for use in ergonomics analysis. However, little has been published regarding the effects of the many decisions that are necessary when using the typical methods based on standard anthropometric dimensions. In this paper, we compared manikins generated using traditional methods with those created from direct analysis of body shape. Boundary manikin analyses do not achieve quantifiable accommodation assessments; statistical population accomodation models or virtual fit testing are necessary. However, boundary manikins are useful for representing range of extreme body dimensions while incorporating covariance. This paper provides some guidance for thinking about the various decisions to be made in generating a boundary manikin family.

Update: 2017-06


Several faculty from the Biosciences Group participated in the Enhanced Safety of Vehicles (ESV) conference in Detroit this month.

Hu, J., Klinich, K., Manary, M.A., Flannagan, C.A.C., Narayanaswamy, P., Reed, M.P., Andreen, M., Neal, M., and Lin, C-H. (2017). Does Unbelted Safety Requirement Affect Protection for Belted Occupants? This article has been accepted for publication in Traffic Injury Prevention.

My colleague Dr. Jingwen Hu gave an overview of a large-scale study funded by NHTSA and conducted collaboratively with General Motors. Under FMVSS 208, automakers are required to certify their vehicles based in part on a frontal crash test with unbelted dummies. Several automakers have petitioned NHTSA to permit them to skip the unbelted test if they have belt assurance systems that can be expected to produce near 100% belt use. In this study, we conducted a large number of simulations to optimize the restraint system with and without the belted test. The major change in the restraint system in the absence of the unbelted test is to remove the knee bolster, which changes load sharing between the remaining components of the frontal protection system (seat, seat belt, steering column, and steering wheel airbag).

Jones, M.L.H., Ebert, S.M., Hu, J., Park, B-K D., and Reed, M.P. (2017). Proximity to the steering wheel for obese drivers. Proc. 25th ESV. Detroit, MI.

steering wheel clearance

Dr. Monica Jones presented the first study to examine the spatial relationship between the steering wheel and torso for individuals with high body mass index (BMI). Her study demonstrated that none of a cohort of 52 obese drivers sat with the recommended 250 mm of clearance to the steering wheel, and some, particularly those with short stature, sat with essentially a contact condition at the lower wheel rim. These findings may have important implications for crash safety, because the restraint systems are tuned for frontal impact conditions in which the airbag is able to deploy fully prior to contacting the occupant. We are conducting simulations studies using morphed parametric human body models of obese occupants to understand how to optimize restraints for these individuals.

Hu, J., Zhang, K., Fanta, A., Hwang, E., and Reed M.P. (2017). Effects of male stature and body shape on thoracic impact response using parametric finite element human modeling. Proc. 25th ESV. Detroit, MI.

morphed models

Over the past several years, our group has developed automated methods for rapidly modifying the size and shape of FE models used for occupant simulations in crashes. This study was the first to morph both widely used detailed models of midsize male occupants, from the THUMS and GHBMC families. The results demonstrate the utility of being able to separate true anatomical effects from the differences between models.

Update: 2017-05

A new paper on our work modeling child body shape has appeared in Traffic Injury Prevention. My colleague Dr. Daniel Park led the development of the first-ever statistical body shape model of seated children. The model, which is available online for interactive use, predicts body shape for children ages 3 to 11 years as a function of stature, body weight (expressed as body mass index), and the ratio of sitting height to stature. The posture of the predicted body shape can also be varied with respect to recline angle and lumbar spine flexion. Whole-body laser scans from 135 children in up to 4 postures were used to create the model. Applications include the development of new physical human surrogates for safety system design (e.g., crash test dummies) and parametric human body models. The downloadable body shapes are also approprite for use in the development of child restraint systems. The new model joins other body shape models for adults and children available online at

Update: 2017-04

The Biosciences Group was well represented at the SAE Congress this month. My colleagues presented three papers on a range of topics.

Jones, M.L.H., Park, J., Ebert, S., Kim, K.H., and Reed, M.P. (2017). Effects of Seat and Sitter Dimensions on Pressure Distribution in Automotive Seats. SAE Technical Paper 2017-01-1390. sipt

My colleague Dr. Monica Jones presented work investigating the associations between sitter and seat characteristics and the pressure distribution at the seat-sitter interface. A study was conducted using 12 production driver seats from passenger vehicles and light trucks. Thirty-eight men and women sat in each seat in a vehicle mockup and seat surface pressure distribution was measured on the seatback and cushion. Anthropometric dimensions were recorded for each participant and standardized dimensions based on SAE J2732 were acquired for each test seat. Regression models were effective in predicting characteristics of pressure distribution from the anthropometric variables and SAE J2732 dimensions.

Park, B-K.D. and Reed, M.P. (2017). Characterizing Vehicle Occupant Body Dimensions and Postures Using a Statistical Body Shape Model. SAE Technical Paper 2017-01-0497. sipt

My colleague Dr. Daniel Park presented a novel method for quantifying vehicle occupant postures and body shapes. The methodology was demonstrated using children and a single Microsoft Kinect sensor. The challenge posed by the noisy and incomplete data was addressed by fitting the data using a statistical body shape model (SBSM). The SBSM used in this work was developed using laser scan data gathered from 147 children with stature ranging from 100 to 160 cm and BMI from 12 to 27 kg/m2 in various sitting postures. A principal component (PC) analysis was conducted based on these scans along with the manually-measured body landmarks, and 100 PC scores were retained to account for 99% of variance in the body shape and sitting postures. A PC-based fast fitting method was applied to estimate the occupant characteristics by fitting the SBSM to an incomplete depth image of a subject. The results demonstrate that a fast, inexpensive system can be used to produce useful estimates of occupant characteristics that could be applied to improve personalization of component adjustments, restraint systems, and infotainment systems.

Hu, J., Orton, N., Gruber, R., Hoover, R., Tribbett, K., Rupp, J.D., Clark, D., Scherer, R., and Reed, M.P. (2017). Development of A New Dynamic Rollover Test Methodology for Heavy Vehicles. SAE Technical Paper 2017-01-1457.

The SAE J2114 dolly rollover test is the most widely used vehicle rollover test procedure. However, it requires the test vehicle to be seated on a dolly with a 23° initial angle, which makes it difficult to test a vehicle over 5,000 kg without a dolly design change, and repeatability is often a concern. My colleague Dr. Jingwen Hu presented a new dynamic rollover test methodology that can be used for evaluating crashworthiness and occupant protection without requiring an initial vehicle angle. A custom cart was designed to carry the test vehicle laterally down a track. The cart incorporates two ramps under the testing vehicle's trailing-side tires. In a test, the cart with the vehicle travels at the desired test speed and is stopped by a track-mounted curb. While the cart is being stopped by two honeycomb blocks, the vehicle slides laterally from the cart with the far-side wheels sliding up the ramps, which generates the desired lateral roll rate. The vehicle near-side wheels slide onto a high-friction surface, which generates an additional strong roll moment around the vehicle center of gravity. Three physical tests using three armored military vehicles were conducted using the procedure. All tests resulted in the desired 5 to 8 quarter-turns of the vehicle, and the instrumented tests showed repeatable initial roll rates. The tests demonstrated that the newly-designed rollover procedure is suitable for vehicles that are generally too large/heavy for other dynamic rollover methods, and may also be useful for lighter vehicles when a well-controlled, directly lateral roll is desired.



Update: 2017-03


We recently completed an interesting study looking at upper-extremity activities in driving. We coded nearly 10k snapshots from videos of over 100 drivers who participated in a naturalistic driving study. The activities of the left and right hand and forearm were documented, including contacts (e.g., steering wheel or armrest) and whether they were holding an object. Drivers had left, right, and both hands on the steering wheel in 64%, 46%, and 28%, respectively, of frames in which the hand placements could be determined. The driver’s left elbow was in contact with the door or armrest in 18% of frames, and the driver’s right elbow was contacting the center console armrest in 29% of frames. Women were more likely to have a phone in their right hands than men, and women were twice as likely as men to be wearing sunglasses during trips taken in daylight hours. For more information, see the report, now available online.

Update: 2017-02


I gave a brief presentation at the Midwest meeting of the American Society of Biomechanics this month. Held at Grand Valley State University in beautiful downtown Grand Rapids, the two-day conference featured keynote speakers and a dual track of primarily student presentations covering a wide range of topics in biomechanics. My talk covered some of our recent work in human body modeling, with a particular emphasis on using Microsoft Kinect to create subject-specific avatars. Software developed by my colleague Dr. Daniel Park fits a statistical body shape model to data from a single Kinect sensor, rapidly generating an accurate representation of body shape even for a clothed scan.

Update: 2017-01


My colleagues Dr. Lauren Zaseck and Dr, Jingwen Hu have led the publication of recent work aimed at improving crash protection for soldiers. Vehicle crashes and rollovers, sometimes secondary to IED blasts, are a leading cause of injury. In this project, funded by the US Army TARDEC through the Automotive Research Center, we conducted a series of sled tests to assess the performance of various restraint systems. Hybrid-III ATDs were used along with a range of body armor and body borne gear configurations. The results demonstrated that advanced belt system features such as pretensioners and load limiters substantially improve restraint performance. The data also showed that achieving good belt or harness restraint performance with soldiers in body armor and wearing gear is challenging.



©2018 Matthew P. Reed and The University of Michigan