I think those of us in manufacturing are really feeling the pinch of the reams of data now surfacing from all of these plugged-in machines and therefore there is a bum rush to staff up in data science … way faster than the market is producing quality graduates.
The role every company had in droves were “business analysts,” but as time went on it was no longer a cool enough moniker to attract talent. As “software ate the world,” the tools also got a lot more sophisticated such that with a reasonable amount of training you could extract sweeping insights from even data swamps, to say nothing of well-ordered data lakes.
The issue has come down to training. If you want to call it data science because it now actually has some real science in it, what is an adequate level of instruction? When we talk to some of the world’s largest manufacturers about this, they have some Jedi Knights of BI and Data Science, but they all acknowledged that at the plant level, the cobbler’s children go barefoot. They wait for access to these folks that never comes, and their list of data analysis needs grow exponentially. Clearly, turning data into information and insight is the new way to converse, improve, and ultimately win in business, whether you make cheese slices or flying cars.
In our desperation facing the absolute dearth of data science resources, we’ve turned to boot camp graduates. In theory it is a great idea, but not unlike Trump University, it is a little hard to tell the real schools from those that may be less reputable. I know of one company that hired a data science boot camp graduate only to learn that the poor guy was still struggling with basic business and data manipulation concepts. Sadly, they would have been better off with the “maids of milking” that today’s installment replaces in the traditional Christmas song. At least they would have gotten something for their investment.
In the manufacturing realm, making sense of entirely new types of data for a new graduate may not produce results that companies are looking for. Clearly, plant managers need access to qualified resources that understand their business AND data science. We are going to need to partner with HR to come up with credible training courses to train people already in the plant. We also are going to need to ensure that the company’s definition of OT includes the tools to build data lakes, manipulate them, mine them for insight, and distribute that insight to the leadership team. Arguably making sense of data is a new core competency everyone needs.
So, boss, we are going to send the data science boot camp graduates to the line to look for misprinted labels to begin learning the business. I am anxious to see what you have in store for me on day five. In the meantime, would you like a glass of milk?
For what it’s worth, at Parsable, we see just how hard it is to tame reams of data from the factory or the field. What we’ve done is make the process of making sense of data easy, from simple approaches for mobile data collection, to real-time data pipelines, to pre-built dashboards for the common processes you need to measure. Sure you definitely need to ramp up, but it shouldn’t take four data science boot camp graduates – or even one, for that matter.