Value of Context

Imagine you’re sitting in your living room watching the evening news. You are waiting for the sports report to come on because you want to know the score of tonight’s game for your favorite team, (any one will do here, but to be centrally correct, let’s just say … ) the Indiana Pacers.

The sports announcer comes on, makes a few introductory remarks and then says, “Now, turning to basketball, our Indiana Pacers scored 97 points tonight in a hard fought battle. The leading scorer had 32 points, 14 rebounds and 7 assists. And in other sports in our city, … .” Would you be satisfied? Did you get the information you wanted? 97 points is pretty good, right? Why are you just about to throw your beer mug through the front of your TV screen? What is missing? Context. You don’t have any and therefore you have no idea whether or not your team did well in tonight’s game. So, specifically, what was missing?

Well, first of all, you don’t know what team they were playing. That might be important. Was it a league leading team, or are they in the cellar? And, it might be nice to know how many points the other team scored – don’t you think? As a matter of fact, the announcer didn’t even mention which team won. Who was the leading scorer and was he on your team … or the other one? These are examples of context, without which you can’t decide whether to be happy or miserable about how your team fared in tonight’s game. You don’t know if you’re one step closer to the playoffs or whether your team is going to start its spring vacation a bit sooner than the rest.

Let’s take this analogy one step further. Suppose you were a person who occasionally makes a gentleman’s wager on things relating to sports. Partway into the season, an acquaintance said he would bet you $X.00 that his team finished ahead of yours at the end of the year. Would you take the bet? What context is missing that would help you to decide whether or not to make that wager? Well, some might just take it no matter what, merely in support of their team, right? Well, if they knew what X was (the amount of the proposed wager), they might. If X=5 or if X=50,000 could be a bit of context that would help in making that decision with, or without, any more information. The greater the value of X here, the more context you might want before making that choice; like, what team is his team and where are they in the standings compared to your team right now? The greater the value of X, the more context you would want – wouldn’t you agree?

Is Context Important in Manufacturing?

Suppose you are the CEO of a global auto OEM manufacturer that provides more than 1,000 parts to twenty auto assemblers around the world. The Hoboken plant manager has called a meeting to introduce you to his line supervisors and they were making a presentation to you about their respective areas. Now suppose the first one stood up and put the following chart up on the screen and asked if there were any questions.

• Would you be satisfied with the presentation?


• How would you feel about how well this particular plant was doing within your manufacturing operation?


• What conclusion can you draw about Area 5? What about Area 5 vs. any other Area, say Area 4?


• Would you have any questions for this particular line supervisor; this plant manager? If so, what might they be?


• What’s missing here? … Context.

• Moreover, suppose this meeting was to request new capital funding to double the production capacity for all six areas. What would be your decision based on this presentation?

The fact is that unless you were intimately familiar with this particular manufacturing plant and these particular manufacturing areas … and what CEO is that well informed these days … you would be completely in the dark here.

Let’s quickly draw up a list of the information you might need to understand what’s going on at this Hoboken plant.

1. What period of time does this chart represent?
2. What products were being produced in each area over this time period?
3. What is the capacity for each area for the products being made?
4. What is the typical productivity for each area, for the products being produced?
5. What were the targets for each of the products in each area during this timeframe?
6. What is the cost to manufacture, per unit, of each of the products being produced in this chart?
7. What is the selling price of each of the parts produced in each area?
8. If any of these values represents a shortfall, what are the root-causes of that shortfall?
9. How flexible are each of these areas – i.e. could they be configured to manufacture any of the products, or is each one pretty much configured to manufacture just the one product?
10. What is the historical productivity of each of these areas for the products represented in this chart?
11. If this is a single shift, how do the other shifts (if any) stack up against this one in terms of productivity?
12. How do the various areas in this plant stack up against similar areas in other manufacturing plants?
13. What are the major items that have impacted productivity: equipment downtime, inventory issues, employee/labor related, shipping constraints, …???
14. And as the King of Siam was fond of saying, Etcetera, Etcetera … Etcetera!

This is Contextual Information ALL Related to the Manufacturing Results in this Chart.

Virtually all of this information can be found in the form of data stored in various systems elsewhere in the manufacturing company – in other control systems, historians, LIMs, equipment logs, view logs, ACCESS Databases and MSSQL Server, MES and enterprise ERP systems; oh, and less we forget in the myriad of EXCEL spread sheets that are stored on individual hard drives throughout the company.

The information that is NOT embodied in those systems is very likely no longer in existence; spilled out onto the floor with the completion of each production run. For example, the time it took to make each unit or lot of units in Area 2 depicted on the chart. This information was once contained in the control system, but it’s unlikely that it has been retained therein – so unless someone recorded those times, or the control system was linked to a logger or external database or historian, the information may no longer exist. Therefore you, the CEO of this large company, cannot find out how long it took to make the 135 units, or whether that time was good or bad relative to any standard, to other units made previously, similar units being made in other areas or plants. You can’t tell if this production area in the Hoboken plant is doing well … or not.

So, Mr. CEO, how does that make you feel – frustrated? … angry? … or is it unimportant to you? If you’re just a “bottom line” kind of guy, wouldn’t you expect your Plant Manager or the Production Manager to know this stuff? Perhaps the detail is too granular for you, but if this visit has taught you anything, knowing whether the Hoboken plant is performing acceptably – along with similar information from the other 40 plants around the world – is fairly important to you. You may even be wondering what it would take to give you this information on a regular basis without having to wait for those end-of-month financials that seem to take so long to get.
You might be surprised to hear that there is software technology from Rockwell Automation (yes, the very same company that provides most of your control systems in those 40 plants) that can deliver that information – to any level of detail that you would like to see – right to the laptop PC on your desk, through a simple Web browser – any time you want to see it. What’s more, this technology is embarrassingly inexpensive compared to the other enterprise technologies your company probably uses right now; and it is in use by major food, pharmaceutical, chemical and energy companies around the world … right now.

This technology is known as FactoryTalk and includes FactoryTalk Historian and FactoryTalk VantagePoint EMI (which stands for Enterprise Manufacturing Intelligence). You can learn more by contacting us, CLICK HERE, or going back to our website.

Marginal Cost of Curiosity

Years ago, as a young engineer working for one of my early mentors in manufacturing, we were experiencing production results that were continually below targets. We would look at the monthly production reports, sales forecasts and shipping details to try to understand why we were under goal again - and we would make some corrections going into the next month. But each month, the story was different - something else had gone wrong and all the corrective action we had taken seemed to have had no effect.

Earl, head of manufacturing, said he needed information from manufacturing sooner and more frequently than once a month. So he set up a team of people to go out onto the floor every two hours and gather a handful of parameters; some about production, some about quality, some about inventory and production kits.

We would compile this information, and four times a day he had a snapshot of what was going on. Nothing seemed to happen. We gave him the report, he quickly scanned it and dropped it on his desk. Then one day while reading one of the reports he said that we were experiencing a yield problem on one of the production lines due to a bottleneck in the test and rework cycle.

For the life of us, we could not see how he could have gleaned that from the report we had provided, but one of the team was sent out to follow up. Sure enough, there was a problem that uncorrected could have resulted in very high cost problems down the line during the system quality testing; or worse, when product had been shipped to customers.

Coincidence or cause-effect?
There were a couple of other, similar incidents through the remainder of the month, all seemingly minor in-and-of-itself. Yet, when the monthly report came out, we had met or exceeded all production goals. Over the ensuing months, we refined the data we gathered and fairly well institutionalized the practice and at the end of the year we had met or exceeded production, quality and resource utilization goals in all but one of the nine months remaining in the year. More importantly, we had met all production goals except asset utilization for the entire year!

Earl was promoted to general management of another division and his replacement felt that the cost of gathering and preparing the information was too costly. It was a down year and some of the team was redirected or released. We were back to the original processes for manufacturing and we never again met a single production goal and Earl's replacement didn't last.

Lesson learned - there is great value in the data that is produced during the manufacturing process. Seemingly unrelated tidbits of information viewed with the keen eye of curiosity often reveal startling insight. If we listen to the wisdom in this insight, we can improve our processes even as they are under way. Twenty years ago, to cost of this kind of curiosity seemed to be too high, but today's information software tools, known as Enterprise Manufacturing Intelligence, take the cost of collecting, correlating and presenting this information down, effectively lowering the marginal cost of curiosity to the point where it really can pay off.

This experience is one of many embodied in the team that developed and now deliver FactoryTalk VantagePoint and FactoryTalk VantagePoint EMI. It is their mission to expose that data through thin-client web browsers so that more and more people can exercise their own curiosity to drive cost savings, increases in productivity, quality and asset use.