Researching How Autonomous Should Work For The Cognitively-Impaired

Tuesday, March 9, 2021

#Functional Safety    #Quality    #Team Integration

We all imagine the happy path. Simple instructions. An autonomous vehicle that arrives upon request, drives itself from our doorstep to our office’s lobby and then either parks itself or journeys to its next rider. No hassles. No worries.

Now imagine it’s an autonomous Uber UBER +2.4% or Lyft LYFT +5.8% vehicle that breaks down unexpectedly en route. Confusing. Certainly manageable with the right app, search engine and a cell connection, but nonetheless stressful.

Now instead imagine you have dementia or some other cognitive-impairment. Even the initial two words of “simple instructions” just became a significant hurdle that engineers have barely considered previously. How did they manage through summoning the vehicle? Or directing to the destination? Even worse: the extreme edge cases such as the vehicle overheating become conundrums that challenge the very definition of “accessible”.

Therein, the United States Department of Transportation (US-DOT) has created the Inclusive Design Challenge, which seeks to make autonomous cars accessible to those with disabilities. There are multiple teams competing against each other; one team consisting of The University of Kansas School of Engineering, The University of Kansas Medical Center and The University of Florida is amongst those that recently won a $300,00 grant from the Challenge and will spend 18 months designing a User Experience for passengers with cognitive disabilities. The final USDOT-selected winners (per a demonstration) shall receive a cash prize up to $1M.

“We are focusing a little bit more on people with dementia or mild cognitive impairments,“ explained Alexandra Kondyli, the project leader and Associate Professor of Civil, Environmental & Architectural Engineering at The University of Kansas. “These people might have issues with identifying automated vehicles – maybe even their own or if they have called an Uber or Lyft – and then figuring out how to get inside, operate it during the ride, and interacting with it in an emergency situation."

When pressed how the experts identified the specific use cases and touchpoints that might be of specific interest, Dr. Abiodun Akinwuntan, the Dean of the School of Health Professions within The University of Kansas Medical Center and a member of the project’s design team, stated “We came about the initial list of challenges based upon our collective experience conducting research and literature searches of driving problems encountered by patients with different neurological conditions. For the patients with mild cognitive impairment and dementia, some of their challenges have been with memory, information processing, contextual awareness and purpose of action.” This might be as simple as remembering where the car was parked akin to the Seinfeld crew getting lost within a parking deck, or as complex as remembering the destination and questioning if the current route is headed in the right direction. “We want to ensure effective usage of the automatic driving system, that the passenger is safe throughout the entire process, and that the cognitive challenge does not inhibit any of the things that they want to do.”

Creating A Draft Design

The goal of the project is not to create a final design, but to propose better solutions that might inform government regulations and assist the automotive industry with design direction. “I believe [the US-DOT’s] goal is to also help navigate the design standards and regulations for highly-automated vehicles,” suggests Kondyli. “Hopefully with this research and other US-DOT efforts, the standards will be guided now to inform future designs.”

“We are designing using the User-Centered Design process,” explained Sanaz Motamedi, a faculty member at The University of Florida who researches and teaches Human Factors amongst other things. “That will use three iterative steps: Understand, Prototype and Test. We obviously have the experts on the research team, but we also have access to users with whom we can test the prototypes and improve them.” The design might include a mobile app for pre-trip planning with possibly a caregiver, it might provide ongoing contextual or location information to both the passenger and caregiver, or it might include in-vehicle assistance. “We are trying to detect and monitor the passenger during the trip, and if any medical emergency happens that we inform the caregiver, the doctor or the emergency contact. In addition, we might provide an agent for the passenger to provide resources or instructions. All of these might be enabled via permissions set before the trip.” The team will have a control group of older users as well as the cognitively-impaired participants so they may understand challenges typical users might face in addition to those amplified by pre-existing conditions.

There is an Advisory Board that has some automotive industry representation, and all teams will have some interactions with industry engineers, but invite manufacturers who might be interested to provide input. “Please engage in the discussion,” asks Kondyli. “Help us create solutions that might see production one day."


Do you need to improve your automotive product development, to increase efficiency, or to comply with ASPICE and Functional Safety? You are at the right place.