Biomechanics expert reveals safety concerns of novel seating positions
Jason R. Kerrigan
Director
University of Virginia's Center for Applied Biomechanics
Recently we chatted with Jason R. Kerrigan, Associate Professor in the University of Virginia (UVA)’s Department of Mechanical and Aerospace Engineering and Director of UVA's Center for Applied Biomechanics. Hear his thoughts on how we can keep passengers of autonomous vehicles safe in new seating positions, as being explored, developed and demonstrated in SAFE-UP demo 1, through Occupant Monitoring system (OMS) and sled testing activities.
Q: What do you believe is the most promising or efficient way to protect AV occupants in novel seating positions?
Adapt current restraint systems to the novel position;
bring the occupant into a protected position prior to a crash;
invent new restraint systems for the novel positions;
something else?
A: So far, it seems that restraint designers have had some success adapting current restraint systems to novel positions. However, the positions that have been studied the most are reclined and rear-facing upright, and those positions are not so different from standard positions in that they don’t seem to require major re-thinking when it comes to how to protect occupants. With that said, if such novel seating positions are very different from current seating positions, like laying down, we will most certainly need to re-invent how we are protecting people and identify all new approaches.
I have seen some systems that try to bring the occupant into a good position prior to crash, and while this may be possible at some point, now, I am concerned that these dramatic high-speed motions subject the occupant to injury risks in the moving that will need much more investigation before they can be considered safe enough for use.
Q: Do you expect any negative impact on occupant restraining from emergency braking or steering, if the crash cannot be avoided? When it comes to how it should be evaluated do you foresee that...
a. It's enough to evaluate the in-crash part but with different initial positions coming from the pre-crash phase either with Human Body Models (HBMs) or Anthropomorphic Test Devices (ATDs)
b. The whole pre-crash and in-crash phase must be investigated using Active HBM?
A: I think that with additional research and development we could consider the use of an Active HBM to predict occupant position and bracing state, which can provide the input for a crash test. However, there are two major challenges with this idea. Firstly, ATDs may not be able to be in a position that realistically mimics this situation (i.e. an active dummy that can generate bracing forces may be required to accurately represent the situation). Secondly, active HBMs will need more development and research to ensure we can trust their predictions in cases we do not have data for.
This question needs more study initially before we can decide how to incorporate it in crash safety testing. There are data now on how occupants move as a result of automatic steering and emergency braking. Those data could be used to start examining the effect of different initial positions with a passive Finite Element Human Body Model (FE HBM) in a crash simulation. However, to evaluate injury risks in that state, we would need to deploy restraint systems to an out-of-position (OOP) occupant, and I do NOT believe that our FE HBMs are sufficiently valid to provide accurate predictions in such OOP cases. So, I would start by collecting post-mortem human subjects (PMHS) data in OOP configurations suggested by volunteer testing to firstly, identify what kinds of injuries could be produced, and secondly, generate some benchmarking data that could be used to evaluate models. From there, I would want to make direct comparisons of model predictions for OOP occupants and dummy predictions - through testing or simulation - for similar OOP configurations to determine to which level we could trust the result of an ATD in this test mode.
Q: How do you see the effect of the self-selected individual human sitting posture (e.g. lower extremity position like cross-legged, or upper body leaning outboard resting) on protecting the occupant in terms of a crash? Is this issue sufficiently addressed in current regulations and consumer testing?
A: Self-selected seating positions will have a dramatic effect on injury risk prediction. However, these kinds of positions or poses have not yet been studied in-depth for upright occupants, so I am not sure there is sufficient interest in studying them for novel seating positions. I do NOT think this issue is addressed in current regulatory or consumer testing. There is a lot of work to be done here, like field data analysis to identify crashes or injuries in cases with those self-selected positions, modelling analysis, comparative analyses of injury risk compared to standard seating positions, etc. I wonder if this problem is not studied so often now because there is still work to be done on ideally-seated occupants. But with more study, one could probably propose some test conditions that could be used to incorporate self-selected postures into regulatory and consumer testing.
———————————-
The University of Virginia is a proud member of the SAFE-UP advisory board.
From the Demo 1 vehicle under construction and a partial result of 3D occupant detection
Part of the use cases collected for processing, including out-of-position occupants
