Page 39 - Mobility Management, July/August 2021

Collision Avoidance
Starting next year (2022) twenty of the largest automakers have agreed to equip all new vehicles with automated emergency braking (AEB). In addition to AEB there are a huge number of options for collision avoidance, automation,
and safety in the automotive market ranging
from Tesla’s AutoPilot to the pedestrian crash prevention technologies on a Nissan Altima. These technologies are amazing, and helpful. Any life saved or collision avoided is considered a positive, but they are far from perfect. In fact, pedestrian crash prevention systems may only be effective about 40% of the time3 and automated emergency braking systems are almost completely ineffective outside of extremely controlled scenarios.4
The required level of performance for these systems has been developed and agreed upon by regulatory bodies, and car manufacturers have little incentive to push for higher levels of safety because consumers have consistently complained at any false-positive braking events by the car.5
The automotive industry has shown that balancing more effective collision avoidance against the annoyance of false positives that stop the vehicle is a challenge.
In many ways, applying this technology to
power wheelchairs is more complex than in the automotive field. Some of the major challenges of collision avoidance in power wheelchairs include:
• There is no exterior surface or roof to place sensors on, so you must place sensors around the user and design the system to look past the user’s lower extremities. When the driver inevitably moves his/ her foot or leg, the sensors may see ‘the obstacle’, which leads to more false positive slowing events.
• The boundaries of the vehicle are constantly changing as the seating system moves, legs elevate, the seat back reclines.
• Users need to maneuver in extremely tight
places, often wanting to touch obstacles with the wheelchair footrests or align the seat to another surface for a transfer, which makes it difficult to tell the difference between a desired contact and a non- desirable “collision.”
• The system must be able to rapidly decelerate the wheelchair:
– which can move at a high rate of speed (6+mph) relative to the proximity of surrounding objects.
– without throwing the user out of the seat or causing discomfort to users with poor trunk control. You can’t assume there is a harness or seatbelt in use.
– when the vehicle itself is inherently unstable under many conditions.
• Current motor controllers for power wheelchairs
do not have encoders and therefore do not provide accurate information about the movement of the wheelchair. When the wheelchair doesn’t know what it is doing accurately it is hard to take any precision action.
• Uncontrolled caster rotation (the “caster flip problem”) can cause several centimeters of unexpected/uncontrolled movement at any moment when turning.
All these challenges, plus the fact that wheelchairs can pivot around their center (zero-point-turn) makes the physics modeling of the wheelchair challenging. In addition, the question “what are standard collision obstacles?”, has not been defined in any wheelchair test methods.
For wheelchairs there are an infinite number of possible collision objects and scenarios out there.
While the challenge is difficult, life in a power wheelchair is a life of bumps and bruises and even collisions. Every collision avoided is one less injury, one less repair, or one less stressful interaction with the world.
3https://newsroom.aaa.com/2019/10/aaa-warns-pedestrian-detection-systems-dont-work-when-needed-most/ 4https://www.caranddriver.com/features/a24511826/safety-features-automatic-braking-system-tested-explained/ 5Obtained from manufacturer based on testing to ISO Standard 7176-3.
5
LUCI® SAFETY REPORT