Man & Machine
Today's focus on automation tends to minimize the importance of people, and several speakers made reference to the human side of manufacturing. While describing the fast-track construction of Pepperidge Farm's Bloomfield, Conn., bakery, this year's Plant of the Year winner, David Watson, vice president of engineering, pointed out that speed in erecting the four walls isn't the issue: rapid ramp-up to optimum production speed is what distinguishes successful projects. "Training the workforce is absolutely critical" if fast track planning is going to yield intended results, Watson observed. Similar cautions peppered speakers' remarks throughout the program, up to and including the final presentation on applying Six Sigma to improve product quality.
"Old software never dies, it just gets uncomfortable," was Thompson's wry take, and constant adjustments must be made or the process improvements are lost. Failure to recognize this phenomenon is particularly pronounced among younger engineers, who have unrealistically high expectations of technology. He suggested periodic upgrades and replacement of out-of-date software are a positive event, not only because of new programs' enhanced functionality but also because of the opportunity to retrain people.
Getchell touched a nerve when he referenced food engineering's rapidly aging "elevator assets," baby boomers with decades of accumulated wisdom and fat retirement accounts waiting to be tapped. The crop of new engineers is flat, "and when they do graduate, they're not going to work for General Mills or Hershey, they're going to Microsoft and Intel," he lamented. "Core competency is what differentiates our technological capability from what we could acquire on the outside," Getchell said, and there is a feverish push for systems that capture the essence of engineering designs for future use before the engineering brain drain hits full force.
Beam me up, ScottyOrganizations like NASA and MIT already have developed the technology needed to deliver advanced skills to low-skill workers on the floor of food plants, according to Dragan Filipovic, associate principal scientist at Kraft Foods' R&D center in Glenview, Ill. An electrical engineer with a background in advanced manufacturing systems, Filipovic has spent the last few years investigating ways to improve operating efficiencies, operator training and changeovers by making affordable diagnostic hardware available to food workers.
"Don't discover boiling water," he advised. "Look at what technology is emerging and what your needs are, and make sure you address them."
To illustrate, Filipovic highlighted the Visually Guided Servicing & Training Project prototype recently unveiled by Kraft. Kraft hopes to deploy VGST through technicians sent to service machines about which they have no knowledge. Wearing customized goggles with a miniature camera and display monitors and carrying 3.2 gigahertz computers on their belts, the technicians will be able to access video clips, PowerPoint graphs and other material about the machine loaded on the computer's database. Using the camera the technician could collaborate with a specialist at the machine manufacturer's location and receive step-by-step repair instructions.
"There is a lot of backlash to this kind of approach because of job security concerns," Filipovic allowed. "It's a helping tool. I would never say to a trained maintenance technician, ‘We don't need you now, you're fired.'" Nonetheless, he predicts "a big, long fight" before this type of technology migrates to the plant floor.
"Technology today is providing all the tools, and it's not that expensive," Filipovic says. "Let's connect the dots."
Avomex operates three plants with about 10 batch and semi-continuous HPP units in Sabinas, Mexico. "Even though we're limited by the batch process, we make it work," she said, with almost 1 million lbs. of avocados alone processed each week.
Grocers such as H-E-B Foods "want high-quality fresh products but don't want to compromise on safety," and they are very receptive to HPP because of its effectiveness against bacteria, viruses, mold, yeast-virtually every foodborne microorganism with the exception of spore formers, she said. Avomex's copacking business is taking off because of HPP's economic barriers. A high-pressure press cost "well over $2 million" a decade ago, Walker noted. Avure Technologies now is selling units in the $1.2 million range, and reliability has improved.
Improved performance with existing equipment is the challenge for most processors. Statistical methodologies to increase throughput and reduce deviation exist, but the food industry generally has done a poor job relating statistical analysis to desired outcomes, observed Monty Vandenberg, director of Six Sigma at Bama Companies in Tulsa, Okla. "In the food industry, statistical analysis often is replaced by deference to (an expert) who knows all the secrets of making a product," he said. Unfortunately, internal experts have different criteria than customers, and statistical measures are useless if they focus on the wrong parameter. The only way to boost customer satisfaction is to first understand their needs, then develop the measurement metrics and process design that will deliver them. Quality control charts and other statistical tools then become the validation tools to ensure the process is delivering them.
Bama's Six Sigma program began three years ago and has resulted in $3.2 million in throughput improvements in the biscuits and pies area and $1.5 million in scrap reduction in frozen dough. Last year, total savings were conservatively estimated at $3.8 million, according to Vandenberg. "Six Sigma isn't necessarily about plant operations, or throughput, or efficiency, or product quality. It's a tool box that can fix anything."
It's also a continuation of a quality focus that includes application of W. Edwards Deming's principles of Total Quality Management. "Deming is seen as a statistical guy," Vandenberg said, "but his biggest emphasis was on psychology."
It was an apt reminder of a point made by one speaker after another. The automated processing facility has a toolbox full of technology to improve product quality and reduce variation, but variation in human behavior is the most difficult to control. Effective plant managers understand that and address both needs with the programs they implement.