The advent of Industrial Internet of Things (IIoT) platforms has made it possible for a proliferation of narrow solutions that target specific and well-defined problems. Most of the available platforms provide a development environment designed to build any kind of application, not necessarily specialized in manufacturing. Part of the value proposition of these platforms is flexibility and fast development. They focus on easy connectivity, easy data transportation, easy cloud storage and fast front-end development, not having any specific module for automation, manufacturing execution system (MES), quality, maintenance or any other typical functionality required in the operations. They leverage the concept that, in any case, a solution based on any of the “traditional” platforms would need to be customized and tailored to the specific needs of the client. So why not just provide a good tool to start from scratch and deliver fast?
I’m not criticizing this approach. During the 20+ years of my career developing software solutions for manufacturing, I’ve always been a strong supporter of tailor-made solutions. The solution needs to adapt to the client environment and culture, and not vice versa. The strength of many companies is not in their physical assets, but in their organizations’ ability to respond to market requirements. Similar companies producing the same product can have totally different organizations, and you need to be able to accommodate that. You need to adapt to it—empower the good habits and maybe correct the bad ones, but you cannot change the way the organization operates without the risk of compromising its efficiency.
So having a tool that enables you to create fast customized solutions can be a good approach. But I see a risk in that also: the risk to try to solve micro problems without having a complete view of the company needs and to build multiple systems that cannot be integrated to provide a comprehensive management tool, or that cannot grow or adapt to the continuously changing company needs.
Fast and free is intriguing, and I listen to many clients consider trying to start solving a small problem without spending any time or money to do an analysis of their real requirements and of what the requirements could be in the near- or long-term future.
There’s value in a fast and cheap solution, but there’s a lot of value in an integrated, scalable solution as well.
When I approach a new client I always tell him: “Spend some time in thinking how you want your company to be in five years, and then start building what you need the first year.”
Moving fast to solve some niche problem—and, by the way, perhaps implementing some great solution from a technological standpoint—can compromise the possibility of reaching the long-term goals. It can at least make it much more difficult. I’ve already seen some clients implementing specialized solutions, or subscribing to specialized services, to solve a problem, and then in a couple years being bound to that solution and unable to expand it or integrate it in a more holistic approach.
Many times, they think that, because it’s cheap, it will be no big deal if they decide to throw it away and substitute it for something else. That might be true for the application, but not for the data. Many companies underestimate how much data and information become an integrated part of the production tools. You can run a factory without them, but when you get used to having them available, you will soon become dependent on them and feel blind or disoriented without them. This is what makes it difficult to change a solution once you have implement it; it can become a significant constraint in moving toward an integrated system.