Introduction

Large production systems have always existed they just haven’t seen themselves as coordinated. Nor have they been measured or tried to perform as a system. Concentration of ownership and the formation of coordinated systems have automatically created new sets of data. Managers and leaders have attempted to assimilate the data into information. Information should lead to the knowledge useful for managing a business.

Measurement without prediction

As mentioned earlier, there is no lack of data and observations in the swine business today. One could easily make the assertion that the data are expensive to collect, hard to assimilate in a form digestible to the workforce, but does do a reasonable job describing what has happened. However, information is rarely organized in a fashion which helps a manager understand much about WHAT MIGHT HAPPEN! At the end of the day, management is about influencing results. In a world laden with so many inputs and elements impacting the process and results of a process, questions come to mind such as, “What should I measure”, “How should I use it?” and “Which results really mean something?”.

What information should I even bother to collect, and how should it be assimilated?

Most are aware of the presence of variation in results. For example, does it surprise us that in week of 11.0 average born alive, there are ‘normally’ sows with born alive numbers below 8 or above 13? I suspect not. How about a series of weekly born alive averages that ranges from10.4 to 11.1? Not really. How about a sow that only has two pigs? ‘Stuff happens’.

The real challenge is knowing when to perform intervention as a result of seeing something in the information. More clearly stated; “Which occurrences of variation can be connected to an identifiable cause?” and “How should I teach people how to separate normal variation from that variation that has a source that is in their area of influence?”.

Deciding what information to get people connected to

In this world of high variability farm staff are often hungry for guidance which will help them understand the information they should focus on. In Murphy Farms, we prefer to take a systems approach. There are some variations in the approach, but in their most basic sense, each of them consider how they will do the following;

1. Define objective

2. Identify drivers/factors

3. Develop and implement methods

4. Measure

5. Refine process

Control charts should be used for the measures (or at a minimum simple run charts) so that individuals in the process can see how normal variation exists. When working to change a process (to try to cause better results) or when “unusual” variation appears, a predictable problem solving process should be undertaken. Any process that produces results should have low variation in how it is performed. Root cause identification is no exception. Helping staff understand the impact inputs have on results, giving them tools to analyze the most important of those inputs and showing them a process on how to measure the result of their effort has organizational power like no other incentives available.

Large systems behave more like the industry

The power of computers has encouraged us to think that understanding more and more about an individual animal will lead the commercial industry to a greater understanding of how the system can be improved. Computing prowess may in fact just lead us to incorrect assumptions much quicker.

Seasonality has always existed in pig production. Simply looking at April and October market hog prices demonstrates the point. Historically we have thought this was driven by seasonal farrowing plans in facilities that were not well enough fortified to protect young pigs from extreme winters. Seasonal behavior still exists even in today’s large systems with carefully managed sow and pig environments. The impact of production seasonality in highly capitalized fixed grow finish space might have an even greater industry impact than it historically has (in terms of price/profitability).

New areas of opportunity

Agriculture is a classic demonstration of chaotic behavior. This is demonstrated even in large, supposedly coordinated systems. However, even in this chaotic atmosphere, dramatic improvements in performance have been made over the past 35 years. Controlled environments, improved nutrition, and medical intervention have changed the cost structure, but hopefully also the predictability with which the industry can deliver performance results. Percent lean in carcasses is a good example. The industry has reduced the variation (increased predictability) through improvements in the inputs (methods, materials, environment).

The next steps of improvement (or differentiation?) for the swine industry may manifest in areas that many don’t even have on the radar screen at the moment. The “quality” of the meat in the animal we produce is a very important attribute. There is no lack of attempt to produce objective measures for benchmarking. (pH, color, WHC, cooking yield, etc). Theories abound regarding the inputs that produce beneficial and detrimental impacts. How does a producer sort out what is needed when the customer cannot define the benefits to his business anywhere but in a lab setting?

The answer is the same as the 5 step process described earlier. Define the Objective, identify drivers and get on the path so people can measure their impact. With all the other opportunities, it becomes difficult to evaluate these emerging factors against what we already know we haven’t gotten done back at the farm.

Other areas?

How do we balance animal care intended to improve productivity with social concerns? How do we measure social attitudes about our industry as we make adjustments in rearing methods as a result of the apparent pressures? Can we make those adjustments AND make improvement in cost/predictability/profitability?

Summary

Computing power has provided the swine industry a very powerful tool. While many recording systems help us to better understand the individual animal (through use of sow cards and treatment records as an example), few of them really help lead us to understand the predictability of a large system and it’s normal seasonal behavior. How we assimilate information can help people lead themselves.