In episode five of #DigitalDialogues, I was joined by my colleagues James Hartwright and Glenn Clarke, as well as Scott Taylor from Port Waratah Coal Services to talk about asset health and the journey to condition-based maintenance. This was a follow-on conversation from episode three, in which we focused on developing a single view of assets. It was great to have the opportunity to go into more detail around condition-based maintenance with James and Scott.
There was one question, in particular, I was looking forward to asking my colleague James Hartwright. Given the speed at which technology has been developing, and the role AI already plays in maintenance today, I was curious to know: Will AI replace human decision making in asset maintenance any time soon?
Here is James’ response to that question.
If you prefer to read, here is a quick summary of James’ answer.
James explained that having the human in the loop is still an essential element of efficient maintenance and decision making. He also went on to explain that the AI that’s currently used in maintenance tools is based on what’s called supervised machine learning. It’s about using the extensive experience and knowledge engineers and maintenance workers have to train machines – and that involves humans checking and, when needed, correcting AI – so humans are still really important for that part.
But even beyond that, James does not think that even the most well-trained AI will be able to identify not only what’s going on, but also understand the context and why something happened. However, that second part is critical to making the right decisions.
In summary, many might be relieved to know that machines are not taking over yet. James is pretty confident that we will need humans in the loop and as the final decision-makers for a very long time. AI will support these humans to do their jobs better, faster and saver, but it will not replace them any time soon.
If you want to learn more about asset health and the journey to condition-based maintenance, watch our full conversation on-demand now.
