Can we get ready for automation by studying non-automated systems?

Richard, thanks for sharing and your insight!? God speed and much success.

2 of 2 In the worst case, we could look at the grounding of the MV Wakashio that is now leaking oil - according to the news report they were trusting that their planned route is safe. But there is the possibility that in the route check errors were accepted and not verified for whatever reason resulting in grounding on a reef. if you go through the #MARSReports you can find many illustrated examples of these mismatches between the mental model constructed by the human operator and the representation of the data and information processing of ship's systems, or even paper, VHF or SafetyNet information. data https://gcaptain.com/large-bulk-carrier-runs-aground-in-mauritius/

1 of 2 There is the question of whether and action that has been done is necessary or effective in how it is done in the standard non-automated practice - or if that practice is in fact a workaround to reach an end state through the only means practical in that setup. While reaching the same end state/ objective is required the automated system is often more effective by following a different process than that of an operator. The hard part becomes bridging the automated processes actions with the expectations of the human operator so that the system's apparent behavior matches the mental model of the operator. without matching the expectations many operators will not use the systems with their embedded safties - or will partially use them and override aspects they don't "trust" - or in a worse case will make assumptions about how the system will react and overtrust.