The world of big data, AI, machine learning, and the Industrial IoT has become extremely convoluted. These innovative new technologies are now marketing buzzwords. What’s more, manufacturers are often being misled, confused, or intentionally tricked about what these technologies can and can’t do for them… now and in the future.
GE and IBM do a ton of marketing for their platforms which raises awareness for everyone in the market but convolutes what is really out there. Manufacturing data is more important than ever, but the tools used to measure, decipher, and analyze it are all mixed up in one big pot.
Below, I’ll give my subjective view of how manufacturers should be navigating this space. I realize that I am the owner of a manufacturing analytics company, so my perspective is potentially compromised. However, I think you’ll see that I am extremely fair in my assessment of the state of manufacturing data and technology.
Manufacturing technology is full of promises
The major players in the manufacturing technology space confuse technology buyers about what to expect from products like Watson and Predix. They want you to believe that this stuff is quick, easy to do, and that Watson can do everything from cook to cure cancer to optimize your production line.
But, the reality is that Watson is a lot of rebranded old tech that does little of what it promises.
The same can be said for GE; although their claims are not as outrageous as IBM’s. I’m not the only one who thinks this, Tom Siebel had similar thoughts in this Forbes article.
All of this technology could be truly revolutionary, and when combined in the right way can do amazing things, but don’t be fooled into thinking that one of these solutions can handle every IoT and Big Data problem. It simply doesn’t work that way, and most operations professionals know this first-hand.
A lot of this is smoke and mirrors
Think about a company like Monsanto which manufactures many different products from seeds to chemicals. They are not going to use the same tools to monitor soil moisture and content on the farm as they will in a factory producing Round-Up.
Yet that is how the marketing departments of the technology companies and many analysts are treating these IoT platforms. Like hucksters with a magical cure to all your company’s ills. Manufacturing professionals know this is a pipe dream, but the C-suite is being sold this stuff by these technology behemoths.
When you pull back the curtains on these different platforms you find a ton of proof of concepts and few production deployments. The gains being reported are coming from one project that shows promise but is not being rolled out full scale.
Social media and powerful marketing at events are leading many to believe this is what the top manufacturers are doing and implementing, but almost none of them are. They have the same problems that medium-sized manufacturers have.
Manufacturers are caught in a platform war
Major technology companies are all trying to win the platform war, trying to be Salesforce.com without creating a great application to build a business or a platform on. Think about the big platforms out there today, AWS, Azure, Google Cloud, Salesforce how did they start?
These platforms started as infrastructure to deliver applications. With Amazon it was amazon.com, Azure was X-Box, Bing, and Hotmail, and Google… that one is obvious.
These are all platform success stories that everyone wants to recreate. The problem is they are thinking about the how and not the why. Amazon created AWS to solve a problem for themselves, they created something great out of need.
Now, technology providers are focused on building a platform and thinking about the problem after the fact. This means doubling down on marketing instead of technology.
Where do manufacturers go from here?
This is where it’s really hard for me to be objective about where I think manufacturers should be looking for solutions to their biggest problems.
Unfortunately, if you’re a manufacturer today, you cannot trust one of these platforms to solve your business challenges. However, if you understand your problems, chances are you employ someone that knows what data could actually help your business lean out, save money, increase OEE, reduce downtime, you name it…
Manufacturers should be pursuing manufacturing data that is relevant and contextual. This means you can actually do something with it that is measurable and attainable. (Hint: this means it has an actual practical application of data beyond a graph showing temperature over time).
My opinion about manufacturing data
I created Mingo because I want to provide value instantly to manufacturers. A solution that lowers the barrier to understanding the complex problems of machine data and machine performance. Something that crosses the lines between the alphabet soup that is manufacturing software.
Principles that will keep you from getting screwed
Whether you’re interested in anything to do with Mingo or not, I have a few guiding principles that I believe could steer just about anyone investing in technology in the right direction…
- Avoid an army of consultants – sorry Deloitte and Accenture you won’t be reselling Mingo. Well you can, but there won’t be a lot of consulting services attached to it.
- Always choose ease of use and simplicity over flexibility
- How easy will it be to implement? How will it affect and interact with other systems?
- Have a plan to demonstrate fast ROI and cost payback that makes sense.
- Make sure there is also some form of the long term value
- Build open systems that break down silos of data between departments like production and maintenance
- Software should be smart and tell you what you need to know and not make a user dig for the answers
As Mingo grows over time, we will check all of these boxes. This is what I’m passionate about, and I’m tired of marketing departments taking advantage of smart manufacturers that are trying to solve a problem.
We are hyper-focused on this one issue, using data from machines in the factory to improve performance. We are not focused on building the next IoT or manufacturing platform, we provide all the data, in-context, without implementations or costly hassles. Because of this, we provide a faster and better return than the platforms ever could at this point.
There are a lot of problems in manufacturing that still require a lot of manual effort. You have to pull data from many different systems to understand what’s happen or solve an issue. Capacity, quality, utilization, and bottleneck identification are good examples. We want to solve these issues with simple software applications.
Where to go from here
This isn’t a sales pitch for Mingo. I really wanted to deliver as much transparency to anyone who would listen about the nature of the manufacturing technology space. I believe in the principles above. I founded a major technology consulting firm that worked with major manufacturers, and I learned so many lessons in that time. Many manufacturers are still learning these lessons the hard way. Hopefully, this article will help.