How to Sell the C-suite on Machine Learning in the IIoT Era
Date Updated: Monday, August 7, 2017
Lots of cool stuff happens in science fiction, doesn’t it? You have Star Wars with its glowing lightsabers, characters getting beamed up in Star Trek and even self-thinking machines in films like I, Robot and Ex Machina. Except, in many ways that last point is not science fiction - it’s reality.
Ok, so this doesn’t necessarily mean robots with advanced emotional intelligence. But it could mean machines with the power to communicate and solve problems without the need for human interaction. This is being driven forward by the growth of IIoT, or the Industrial Internet of Things. The IIoT will mean that machines can monitor themselves and report issues, long before they become serious problems. ‘Failure of critical assets’ was rated as the most significant risk to operational performance in a recent survey, as almost 40% of executives considered this as having the greatest impact on industry operations.
A foolproof asset failure prediction plan not only carries the ability to maximize the reliability and availability of factory assets and reduce unintended downtime, it can also boost equipment uptime by enabling early detection and prevention of the failure in an automated environment. As leaders of enterprise industrial organizations, the C-suite's priority is to get and empower their field technicians with an accurate visibility into what does all their asset data really translate into, which assets are most critical to meet business objectives and most importantly how can their entire organization march towards the ideal scenario of zero outages.
In our experience, companies truly like to seek out for answers to questions to slice down impending asset failures which aren’t open-ended but very specific. For example,
- What? Which is the root cause of recurring asset failures?
- Why? Is there a history of asset failures / does a failure pattern exist?
- How? Can you predict the actual cause of the failure?
- Who? Who are the asset operators responsible for performing maintenance tasks?
- When? When can you achieve the optimum asset efficiency?
- How much? How often is a maintenance intervention required?
As it happens, a key reason for this growth is the desire to implement predictive maintenance functions within industrial equipment to monitor its health. This helps avoid unscheduled down times in the production cycle, saving money in the long term. By 2022, the market worth of IIoT is expected to reach $195 billion. Yet, even with all these predictions, getting C-suite leaders onboard can be difficult.Nobody Likes Change
One of the problems with the implementation of new technology is that it is disruptive and can also be uncomfortable. It can upend processes, destroy hierarchies and rebuild the very formation of a company. For this reason, existing workers can be resistant to new processes. If there is a risk that implementation could damage company profitability, c-suite leaders will be wary of shareholder interests.
The best solution is to start small. Pilot programs can win teams over by demonstrating the value of new technology to a business’ success. This is called a ‘Proof Of Concept,’ or POC. Many of these initiatives fail because they run over budget, or over time, and fail to properly demonstrate the tremendous value afforded to businesses that Big Data brings from connected machines.
Successful POCs need to be able to demonstrate visualization and analysis, Big Data management, and the storage of the harvested data. Performing a small-scale introduction like this is difficult, but it can help get employees and C-suite members behind new projects and technology if it can properly demonstrate the value it brings.Communication Station
If you’ve ever tried to communicate with someone you don’t share a language with, you know how difficult it is to explain things through a language barrier. Well, here’s news for you - to the uninitiated, new technology (including Big Data, Machine Learning, Data Science) is like a whole other language.
The best way to get around this is to ramp up knowledge and - if possible - enthusiasm before rolling out any new projects. In the case of the IIoT, the top three roadblocks to success are funding, building a business case (see above) and, crucially, understanding. Not ensuring that understanding is comprehensive across the (literal) board, is a mistake that could be the death of your implementation plans. Data ownership is also very important at this stage - facts and figures speak louder than any jargon-loaded dialogue project managers might, however unwillingly, be using.Trust Issues
People don’t trust what they don’t know - that’s as basic as the sky being blue and gravity causing things to fall to the ground. When trains were first introduced in Britain, photos of the then Queen Victoria on locomotives were used to encourage an unwilling public to use them. Similarly, our catch all term for people that hate technology, Luddites, comes from rioting workers in pre-industrial England that resisted the introduction of machines to factories. As you can see, successful implementation is based on trust.
Spending time among workers is crucial to winning over all crush. As a leader, it’s important to come across as a person, not a revolutionary force that will crush their roles and force upheaval upon their daily routines. Spending time with groups, or working in one-on-one sessions, is crucial to gaining trust, and then enthusiasm behind a project. Happy employees often lead to happy bosses and this is a good way of winning over skeptical c-suite members, too.
By following these steps, it is possible to demonstrate the benefits of change, get everyone on board with the language of new technologies like machine learning and data science, and win over the trust of a work force - and c-suite members - to see the positives of the development. This way it is possible to win the day for technology, and your business.