By Dr Nathan G.E. Kirchner, Founder and Chief Science Officer, work_r; Founding Director, Robotics Australia Group
We are in interesting times. We have the technology, we have the means, we have the need, and we definitely have deep pockets of desire, so why aren’t the robots here already? Why has adoption, especially in the heavy industries, been so sluggish?
A slow adoption of robots comes down to three possibly surprising and considerably more physiological perspectives than technical core pillars. Left unaddressed, these cause all sorts of good projects to fail. Thankfully, once they have been highlighted, they are relatively addressable.
First, however, let’s do a quick refresher of the underlying motivation to introduce robotics in the first instance; it’s simple to explain in the context of heavy industries. Plenty of jobs are sufficiently dirty, dull and/or dangerous – particularly in the mining industry – that people simply cannot be asked to do them. Alongside this, between the tyranny of distance and the skills shortage, there isn’t someone to ask. Nevertheless, the work still needs to be done, so bring in the robots!
The challenge isn’t technical
So, why isn’t there more adoption of robotics? Surprisingly, the challenge isn’t technical; it comes down to a fundamental misalignment of what resonates with one group within an organisation, versus another. This typically results in frustration as we talk past each other and fail to build the conviction required to progress.
Let me give an overly dramatic example to simplify my point. Let’s say an office-based group resonates with a piece of functional innovation – this innovation clearly has impact on the business that is enabled by technologies; however, consider a group of frontline people – they are more distant from what the innovation would do for them, and may consider any burden they must ware to adopt the innovation as a cost they pay for someone else’s benefit.
This is, of course, just one example; but consider Figure 1 and your own personal experiences. There are many possible combinations where various stakeholder groups may resonate strongly, while others feel left out.
The solution could be to balance all innovations so that there is ‘something for everyone’; however, it may simply be the case that effective communication can resolve any imbalance and achieve an innovation effort that is widely supported.
We are not 100 per cent accurate
This is just the first hurdle. From here, we generally find ourselves faced with perplexing quagmire. If a person performs a task with 95 per cent accuracy, then we, as people, tend to line up to congratulate them on an outstanding effort; however, if a robot performs a task with the same 95 per cent accuracy at the same time, at the same place, and in front of the same people, we then line up to critise the ‘faulty machine’ for making ‘so many errors’.
This is a fascinating situation, and one that we really need to consider when we are driving innovation into application. Are we trying to outperform the current approach, or are we simply trying to achieve the current level of performance via different means? It will be the later in at least some cases.
If we make clear what we are trying to do, what we are trying to achieve and how we should be measured, we can avoid talking past each other. We are not 100 per cent accurate at what we do, and that’s okay.
Low-hanging fruit poisoned
The starting line is in sight (and onsite), so bring out the engineers! It should be relatively straightforward from here to adoption, right? Shockingly, no. All too often, the low‑hanging fruit is poisoned.
We enter this stage with ambitions of unlocking impact with our innovation – but, all too often, out comes the value engineering. Often, costs stripping becomes the sole focus, and various elements are removed and reshaped until we are left with an anaemic carcass.
We hear triumphant claims of lean design/engineering, and just how much money was saved. Unfortunately, the reality is that the ability of the innovation to have any measurables is frequently undermined. We are less with a cost to test ‘nothing’.
This is extremely dangerous for adoption, as the demonstrated impact is attached to the innovation, whereas it was the test design that hampered the innovation’s ability to impact, rather than it having been the innovation itself. This very sticky misattribution has sank many a promising innovation. We must be cognisant of it, guard against it, and be willing to stop an unfair trial taking place in the first instance, even if it is at the cost of the innovation not happening.
In short
It requires considerable expertise in robotics, artificial intelligence, and innovation psychology to impact the wider business in any sort of deep and flexible manner. We must address this to enable mass adoption of robotics, and to reap the benefits we can already see. How good is your great idea if everyone refuses to use it?
