# Why Big Data and Baseball Aren’t So Different

Digital technology creates, records and stores a lot of data. That data could be from your personal vehicle, your company website or your factory floor.  If it is instrumented, it is churning out data. Status and history comprise the bulk of Big Data facts and that data can inform and guide your everyday decision making.

Most Big Data has one thing in common: It is a parameter measured over time, creating histograms and timelines of events and measurements. This means most Big Data can be looked at as trends. Trends can appear as normal and abnormal, resulting in patterns that distill Big Data into easily understood results. For example, in the factory, you can measure and record:

• Pieces per second
• Temperature at five second intervals
• Pressure at injection

But recording Big Data is not enough. You must harness Big Data to make it work for you. Baseball statistics are an excellent example of harnessing Big Data. In major league baseball, there are an average of 15 games a day or 144 innings a day, 1000 at bats a day and 4400 pitches a day, for 180 days a year. But for an interested fan, data from all those games is not that hard to keep track of and understand. Why? The Big Data has been recorded by the official scorekeeper and reduced to dashboards by the sports industry. Line scores, box scores, records and standings are all presented as concise and familiar charts.

Imagine your favorite baseball team. Their record is 42-42. Their last ten games, they are 8-2. Big Data shows that they are playing better. Their batting average for the year is .253 and their batting average the last two weeks is .295. Big Data says they are hitting better.  The combined winning percentage of their recent opponents is .401. Big Data analysis concludes that they are winning because they are hitting better against weaker opponents.

## Big Data and Manufacturing

This same process of using familiar charts to analyze statistics can apply to the Big Data generated by your factory. Your machinery is instrumented to gather multiple events and parameters. Your on-machine devices (such as PLCs and sensors) relay that information to your manufacturing execution system (MES) where it is stored. The last mile? Your Business Intelligence (BI) tools read the data in real time and present it as dashboards that:

• Confirm the expected operation of your factory
• Shows processess that are headed out of bounds and automatically alert you if action is required
• Highlight problems that require immediate intervention

Let’s carry the baseball analogy one step further: Imagine you are not just a fan, but the manager of the team. You not only want to know what is happening, but what you can do to improve the situation. This is where the value of Big Data combined with an MES solution really come into play. Do you have the ability to implement this situation in your manufacturing plant?

Machine #411 is running slow, delaying its ability to pick up Job 8091. But machine #372 is idle until 3 PM and has the same capability as machine #411. The job can run in 5 hours. It is 10 AM.  Automatically reroute Job 8091 to machine #411 and then get tech over to machine #372 to resolve its performance issue.

This is Big Data and MES at work. Shop floor operations managers, like baseball managers, have to react in real time so that late in a close game, if the competition brings in a lefty pitcher, you can pinch hit a right hander. Automation and Big Data make quick decisions practical and possible.

Plant managers often serve a more strategic role. They ask questions such as, how did we do last month or last year? How can I configure our factory to increase output 30 percent without adding more equipment?  They need to know that machine #186 ran only 70 percent of the time last year and that machine #202 was 150 percent more efficient when running long batches. Or that machine #98 had a maintenance issue every 11 days last quarter.

The plant manager has to see the big picture and put the factory in a position to overachieve last year’s output. He needs the performance and OEE of every work center at his fingertips to makes adjustments and upgrades. You might call it Billy Ball – Getting the most for the least cost and seeing behind the headlines into the work center’s true production.

The CEO plays the “game” at yet another level. They look at data such as:

• My cost of goods is 5 percent higher this year than last
• The average age of my equipment is 9 years
• I have excess thermoforming equipment
• My stampers are running at 95 percent of capacity and my orders for stamping are growing
• We are pursuing more medical device and automotive business and those customers require strict reporting and compliance practices.

With that information, they can make decisions such as: If we re-equip for stamping, sell off thermoforming machines #41 through #63 and win the Toyota job, can we meeting the demand and compliance requirements?

No small task to decide, but that is where the true power of Big Data, MES, ERP and BI come into play.  The shop floor automation records the reality (data) on the floor, the MES collects and stores the data, the BI reports the data in formats that make it quickly understandable and the ERP software lets you run the financial what-ifs that ultimately guide the decision making.

So is manufacturing really like baseball? Of course not entirely, but in many ways the data gathering, presentation and decision making is. Each key player has a need for certain information at certain times and the data needs to be accurately summarized and presented in a usable format. The linchpin lies in the tools (ERP, MES and BI) spanning your entire operation, so your data is integrated and you can work as a team.

#### Steve Bieszczat

Steve Bieszczat, DELMIAworks (IQMS) Chief Marketing Officer, is responsible for all aspects of DELMIAworks' (IQMS) brand management, demand generation, and product marketing. Prior to DELMIAworks (IQMS), Steve held senior marketing roles at ERP companies Epicor, Activant and CCI-Triad. Steve holds an engineering degree from the University of Kansas and an MBA from Rockhurst University.