In our final part of this three-part blog on the recently published Publicly Available Specification ISO/PAS 21448:2019 ‘Road vehicles – Safety of the intended functionality’ (SOTIF) we focus on the topic of how SOTIF approaches human factors and compare this with other industries and their approach to the subject.

 

The term ‘reasonably foreseeable misuse ’is one of the key definitions in SOTIF equally so several references are made to human-machine interface, but ultimately the only part of the specification that touches on how to approach human factors or usability engineering is Annex E – Method for deriving SOTIF misuse scenarios. This three-page annex gives a high-level overview of a human factors process for evaluating potential misuse scenarios. The subject is however, much larger than represented in SOTIF. Industries such as nuclear, medical and aviation put far more emphasis on human factors, and this is an area in automotive that should be given far more attention than it currently receives.

In IEC TR 62366-2:2016 Medical devices – guidance on the application of usability engineering to medical devices, there is an excellent annex on usability engineering methods and the life-cycle phases how and when these should be applied. A total of 21 different methods are described and compared. How each of these methods can be used in hazard, formative or summative evaluations, is well defined. Figure 1 shows how typical methods could be used in SOTIF and at which particular life-cycle phase.

 

Figure 1 Human factors methods at different life-cycle stages

 

Perception, cognition and action (PCA) analysis is represented in Table E.2 in SOTIF, or in this case recognition, judgement and action, however a greater explanation of how this method can be used to identify use errors based on task analysis would be very helpful.

One very positive aspect of human factors engineering is the ‘agile’ approach taken from the very outset of a project, to iteratively improve the human factors aspects of the design. Considerations of demographics, uses cases and environments are given a high priority as the design goes through multiple formative evaluations. The strategies and methods to apply and when to apply them should really be given their own section in SOTIF, to help automotive teams better understand the challenges.

As mentioned in the first to parts of this blog ISO/PAS 21448:2019 is a good overview of the challenges in SOTIF, but there is room for a little more supporting information.

We will be presenting SOTIF the Human Factor at the VDA conference in Potsdam 27-28 June 2019.

 

By Alastair Walker, Functional Safety Consultant