The Fourth Industrial Revolution and its underlying principle of automation was a big-ticket topic in 2017 and will probably continue to be in 2018. With both logistics operators and vendors weighing in on the hotly debated topic throughout 2017 I thought it was worth checking in on automation.
The State of Automation
“Automation has failed to deliver what has been promised by vendors”, commented one senior operator of a global logistics company. They wished to remain anonymous for this article. To demonstrate this with real numbers, they shared data from the 15% of their operations that are, in some form, automated. In short, operations that are automated have seen, on average, a 34% decrease in productivity.
Wait a second. Automation was supposed to offer higher productivity – the exact opposite of what that operator is seeing; right? Well, it turns out, automation is harder than some vendors thought; logistics is complex, so automating logistics is complex. But, not all operators would agree with that. To paraphrase the senior logistics operator, an A380 is complex. Hundreds of thousands of small parts (4 million to be a bit more precise) coming together to perform as a single aircraft that safely transports thousands of people daily is complex. Logistics operations – in absolute terms – is not.
In a discussion with multiple logistics industry hardware and software vendors, there was certainly an underlying message of complexity. From dirty data and resistance to change to unknown unknowns, vendors weren’t shy to say there is a lot in the technology world that is far from perfect. When pressed, every vendor confirmed that, yes, they believe that logistics processes and environments are complex.
Automation in the Future
So where does that leave the logistics industry? Operators have an invested interest in seeing automation reach its potential, as do vendors. However, if operators believe it should be simple and vendors – tasked with delivering it – are saying it isn’t, where’s the middle ground and what does that mean for the future of automation?
The aforementioned senior logistics operator made fair points in the argument toward simplicity. Hardware and software often come with many features that add complexity. With automation, these extras provide one more point of failure that factor into the mean-time between failures calculation, and no surprise, the greater the complexity, the greater the risk of failure. He also noted that, as of today, exception handling within automated processes isn’t done well – after all, robotic equipment only does what it’s programmed to do.
So, if the core of the problem is complexity, why is complexity there; are vendors to blame? Firstly, complexity is often built into hardware and software systems as a result of the attempt to cater for as many known variables as possible. Second, poor design happens and vendors occasionally make mistakes; these things happen, vendors who suggest otherwise are lying. Both of these reasons are realities we all must face. Vendors know this, and as a result, most offer ongoing support programs to mitigate and resolve these issues.
Alternatively, are operators wrong; is complexity an inherit element in logistics? I’m not nearly qualified to answer that, but what I can say is that there are few vendors building hardware or software systems without substantial input from logistics operators. In the eight years (not long, I know) I’ve worked in the IT world, I’ve participated in more tenders and sales processes then I care to count. I can’t recall one where the operator didn’t want a seemingly endless list of features, bells, and whistles to cater for any situation they might come across despite the likelihood of occurrence or the relevance to the business objectives. These sorts of functional specifications are what most enterprise IT software is developed from. It’s no surprise that the resulting systems are complex and do way more than what is needed for normal operations.
A Paradigm Shift is Needed
So how does the logistics industry move forward? Collaboration. Not just between operators and individual vendors, but amongst all industry stakeholders. The more we work together to solve issues with simple, elegant solutions, the less it will cost all parties, the quicker it will be, and the better the end result. This requires a different paradigm for looking at the competition (operator to operator or vendor to vendor) and it requires a different business model for business contracts (operator to vendor) and this is not to be confused with collusion or anti-competitive behaviors.
To start, all parties in the logistics industry need to realize that we operate in the same industry. Working together, sharing resources like data and lessons from our mistakes is a crucial step in improving as a whole. The more siloed we are, the weaker we are as an industry. From a technology perspective, the development of a single data standard and interfacing protocol would go a long way to enabling all stakeholders to more easily share data and derive insights as a whole.
From there, business models must evolve from fixed cost contracts where risk is shifted to the opposite party as much as possible. Traditionally, this has made good business sense, but as noted herein, it isn’t working. It leads to endless conversations about scope creep and change orders that eventually ends with a solution that doesn’t address the real issues – only what was able to be documented at the beginning (known knowns) and not what wasn’t known (unknown unknowns). A business model for the future must be one where operators and vendors share risk equally. With skin in the game from both parties, everyone has an invested interest to see the best outcome.
Quick iterations, quick deployment, and quick feedback enable a continuous improvement process leading to better overall solutions. Fast feedback loops enable operators to test better ways of doing things which in turn improve processes. Enterprise IT developers have largely resisted this framework; why? Risk (see above). The nature of the model means that some percentage of what you try won’t work as expected. So far, the appetite for risk is low but the potential for reward, nevertheless, remains high.
Bringing it Together
Imagine if we knew ten years ago that some automation wasn’t going to deliver as promised. Would we have continued to invest in that technology or stayed the course? Maybe. Could we have changed direction slightly to end up in a different place altogether? Definitely. With an alternative way of approaching competition and business contracts, we’d be in a very different place today.
The advantage of quickly understanding if something will work as expected outweighs the disadvantage of perceived risk. After all, waiting years to the conclusion of a traditional IT project to find out it hasn’t worked as required, is far riskier than understanding the risks up front. As an industry, we must evolve if we’re to reap the rewards of any new technology, because when it comes to technology, the low hanging fruit is mostly picked and what is left, is the hard to reach stuff. This will take time, money, and teamwork to harvest.
Do you have an example of a project that would benefit from a paradigm shift?