Problem Solving and Big Data
Big Data has become a big term these days. Big Data refers to the ginormous amount of data that is available in just about every industry and company. Computing power has made it possible to collect, store and retrieve this data so it can be applied to improve effectiveness and efficiency. When it comes to problem solving, Big Data is a double-edged sword.
Two years ago we met with an executive responsible for production facilities around the world. When we asked about problem solving in his operations, he very proudly demonstrated how he could view the real-time performance of any machine anywhere in the company’s worldwide operations right on his Blackberry. It was very impressive. We said: “That’s great! That can tell you there is a problem, but how do you solve it?”. After he stared blankly at us for a minute, he admitted he wasn’t sure.
Our experience with this executive is not unusual and has become even more common as Big Data has expanded. Some of this data can tell you there is a problem, by which we mean that a machine or process is not producing the quantity and/or quality of product that it was producing previously. In this regard, the data is very helpful in quickly identifying that the machine or process is not doing what it should be doing. In the past, an entire run of product might have been produced before it was determined there was a problem.
But some people think that this abundance of data will automatically fix the problem. It won’t because the data will tell you “what” is or is not happening, but it won’t tell you “why” it is or is not happening. Effective problem solvers leverage data by asking critical questions about the problem. We like to say that effective questioning about the problem converts data into useful information.
When we train employees in our manufacturing clients to use our problem-prevention and problem-solving processes, we have them bring real job issues to the workshop. When it comes time to work on these issues (application of what they are learning) they usually have to get data about the problem. The questioning process we give them pinpoints the type of data they need and helps them organize the data so they can start narrowing down to the root cause of the problem.
Data is the lifeblood of effective problem solving. Without it you really don’t know much about the problem, why it is happening and how to fix it. This “wing it” approach leads to making changes that don’t fix the problem, make the situation worse or cause new problems that are more expensive and difficult to fix. But not having a way to organize the available data can be almost as damaging because you get a false sense of security about the fix for the problem. Occasionally we’ll have machine operators tell us they looked at the data and have identified the root cause of the problem. What they have usually done is better define the symptoms. Some simple questions such as “why this machine and not the others?” or “why now and not before?” quickly deflate their balloon, cause them to dig a little deeper and get to the root cause.
The flip side of the Big Data coin is feeling overwhelmed by all the data that is available. There is a tendency to throw up your hands and quit in frustration. Again, effective questioning can help cut through the data avalanche and focus on what it important. This separating process will leave you with the kernels of information that will answer the question: Why is this problem happening?