Food engineers from across North America convened May 20-23 at Arlington Heights, Ill. for PLANTtech 2001, Food Engineering's fifth annual conference addressing manufacturing issues. PLANTtech delivered a full menu ranging from "e-manufacturing" to ergonomics, from sanitation to supply-chain integration, from extended shelf life to energy management, and featured "case-histories" in plant automation.
What does 'e' mean to me?
Ian D. Steele, engineering manager at Anchor Products, Ltd. (Hamilton, New Zealand), keynoted the conference with an overview of e-manufacturing at Anchor, New Zealand's largest dairy cooperative with 11 manufacturing plants processing more than 37 million litres of milk per day into cheese and dairy ingredients for export. At Anchor, said Steele, "e" means:
- "Embracing" a real-time enterprise that adds value, not cost; continuously adds profit; meets changing consumer demands; delivers continuous improvement, as well as a real-time solution reflecting strategic and tactical decisions.
- "Enabled" technologies that are based on standards and that incorporate neural-model predictive control to optimize product quality, predict performance and analyze failures; produce key performance indicators (KPIs) to benchmark operations in real time; and integrate via the Web.
- "Enhanced" and sustained profitability by focusing on value-added efficiency and flexibility; challenging traditional linkages; re-engineering systems and practices; creating a culture for change; leveraging technology to deliver real-time business value; and optimizing all manufacturing and business functions to process capabilities in order to reflect real-time conditions.
- "Empowered" people trained to manage the process and add value through new technologies, who take ownership and have access to KPIs that reflect how well they're contributing to business and process objectives.
- Enterprise" optimization from process to customer, with rapid proactive response from a real-time delivery perspective.
Challenging the typical "top-down" IT perspective, Steele advocated a "bottom-up" paradigm biased toward production rather than accounting; which reflects dynamic rather than historical business changes; allows real-time analysis and decisions by the minute, hour and day; is based on process models to reduce product variability and on business models rather than assumptions; is proactive rather than reactive; and reflects what is, not what was.
Anchor has partnered with Rockwell Automation (Milwaukee) and Pavilion Technologies (Austin, Tex.) to optimize its enterprise in real time.
Ergonomic training reduces MSDs
"We want to take a proactive role in reducing injuries," said Joanne Breuer, safety manager and ergonomic committee chairperson at the Tillamook County Creamery Association (TCCA), where 500 people produced 61 million lbs. of cheese in 2000. Proactivity is supported with a written ergonomic program; management leadership; employee participation; job-hazard analyses; and training.The Ergonomic Committee focuses on TCCA's 110-person packaging department, where all musculo-skeletal disorders (MSDs) occur. Committee members include the packaging-department manager and assistant manager, three packaging-team leaders, and four line operators.
Assisted by a consultant, committee members first received three days of intensive ergonomics training, then constructed an ergonomic hazard analysis to identify and assess risk factors on every packaging job. Employees completed questionnaires reporting pain resulting from any task, and soliciting suggestions on how to improve the job. Each job is videotaped, and employees critique the tapes.
"We struggled initially" because employees were not involved in ergonomics training, Breuer continued. Last year, however, packaging lines were shut down for an hour at a time so employees could be educated in ergonomics, how it will benefit them, "and this was the turning point," said Breuer. In the three years since its establishment, TCCA's ergonomic program is showing results: Annual MSD claims have declined from 12 to 2, and annual costs from $89,000 to $1,100.
Training for sanitation
"Food safety is unachievable without sanitation systems that employees apply and that are constantly improving," said Dr. Gene W. Bartholomew, corporate manager of HACCP and regulatory affairs at John Morrell &Co.Sanitation training consists of centrally-managed modules targeting training needs in the company's seven plants and are accessible by employees over the company intranet. Training is documented and can be refreshed as needed. Non-professional trainers are first themselves trained in adult-learning concepts, speaking skills, audio-visual aids, action planning, behavior profiles and cultural diversity.
Trainers teach according to a curriculum of modules covering HACCP principles, hazard management, SSOPs, hygiene (customized to each plant), cross contamination, document management, production control and "much more." Documentation includes daily sanitation records, monthly GMP self-assessments, microbiological monitoring of environment and products, corporate-level reviews and audits, and third-party audits.
Software validation
Dr. John W. Larkin, who heads the Hazard Analysis Branch of FDA's Food Processing & Packaging Division, addressed validation of computerized process controls to comply with current and forthcoming FDA regulations and guidelines.System validation can only provide reasonable assurance of correctness, Larkin pointed out. Some excellent software-validation standards exist, however, including those published by the Institute of Electrical & Electronic Engineers (IEEE), the National Food Processors Association (NFPA Bulletin 43-L, currently being redrafted), and ISO-900-3. FDA's general software-validation principles can be found at www.fda.gov/cdrh/comp/swareval.html
FDA is enforcing regulations specific to computerized systems, including 21 CFR Parts 120 (juice) and 123 (seafood), said Larkin. Although 21 CFR 113 and 114 (low-acid canned foods) do not specifically address computerized systems, FDA requires that the record -- including computer-generated records -- be created at the time of observation. For software used to establish commercial sterility, FDA is of the opinion that the calculation shall be performed according to procedures recognized by competent process authorities, and that the software must be validated. A proposed change to Part 106, relating to infant formulas, would require validation of automated systems.
FDA is now developing guidance documents on how it will enforce 21 CFR Part 11 ( Electronic Records and Signatures), Larkin continued. More than 200 FDA inspectors are currently being trained in techniques for inspecting computerized record-keeping systems and evaluation compliance with Part 11. He expects the document to be published in the Federal Register within the next 18 months.
In-line QC
Michael Reid, operations manager for the Flav-O-Rich Dairy unit of Suiza Foods, discussed how in-line standardization of milkfat saves time and money while allowing more consistent milk products at his company's plant in Wilkesboro, NC. The plant processes 17 million gal. of milk per year for foodservice, institutional, school and convenience-store products. After separation of cream, a Tetra Alfast system re-blends fat with skim milk to achieve the desired fat content in whole milk (3.25 percent or more); reduced-fat milk (1.5 or 2 percent); light milk (0.5 or 1 percent); skim milk (less than 0.49 percent) and fat-free skim milk (0.1 percent or less).
Alfast saves time, tanks, product loss and results in a more consistent level of fat in the product. "There's no shut-down between milk runs," said Reed. The system also concentrates milkfat -- sold as ingredients for products such as ice cream, half-and-half, sour cream, whipping cream and butter -- to 43 percent and higher. By adding a flow meter and variable-speed pump "we can blend chocolate milk on the go," he added.
Predictive/adaptive control
Joshua Mensinger, process engineer at the Lonza, Inc. plant in Williamsport, Penn., described model-based adaptive control of batch processes producing food ingredients such as emulsifiers, monoglycerides and sorbitol. The reaction involves long dead times and dynamics change as batch mass increases, making the process nearly impossible to control with conventional PID loops. To improve control, PID loops were replaced with Universal Dynamics' Brainwave control system. Brainwave models and adapts to process dead time, "learns" the process and continuously creates feed-forward models which adapt to changing dynamics, said Mensinger. At Lonza, Brainwave runs on a Windows-NT PC integrated with the existing Rosemount R3 distributed process-control system via an OPC server. Initial models were developed from historical data, then Brainwave was shifted to automatic mode and completed an entire batch on its first run. The system was installed and operational in one week. Results, Mensinger reported: temperature swings reduced from 20 to 5 degrees C; average flow rate increased by 30 percent; reaction times reduced by 20 percent. Payback was achieved in four to five months.MES/batch boosts flexibility
Jerry Leuthold, programming manager at systems integrator Bachelor Controls, Inc. (BCI, Sabetha, Kan.), reviewed an MES and batch-control system customized to govern a dry-mix batter process at the Lamb-Weston plant in Pasco, Wash. The MES integrates a bar-coding system, an S88-compliant Rockwell RSView HMI and associated servers with Lamb-Weston's host business system to control an automated AZO Componenter batching system, blending of the batched ingredients, and dispensing of the dry mixes into supersacks. The MES incorporates a BCI Scheduler which schedules "campaigns" (a group of batches), is formula-driven, and reports lot tracking and ingredient usage back to the business system. "Lamb-Weston's business system creates and modifies the formulas that drive the batch, then the scheduler in RSBatch takes care of the rest," said Leuthold. The bar-code system identifies ingredients for silo filling, Componenter filling, inventory management and tote management. RSBatch controls PLCs, executes recipes, allocates equipment and batching according to the S88 standard, and archives batches. System configuration allows weighing two batches as two more are being mixed, boosting flexibility and throughput. Batch accuracy is within one-tenth of 1 percent.Outsourced energy management
"Enron does energy; Ocean Spray does cranberries!" said Tripta Sarin, vice-president/operations at Ocean Spray Cranberries, Inc. (OSC), allowing OSC to focus resources on its core business and margin enhancement. OSC's 10-year, $116-million partnership with Enron Corp. for total energy management covers 12 locations in seven states and includes commodity management (i,e., negotiating the best rates); information management (one consolidated billing); capital-project management for energy demands; capital and financing services; and consulting services. A multidisciplinary team from both companies defined contract scope; solicited input and support from the manufacturing plants; defined service responsibilities; and developed a timetable for services startup and guaranteed savings. Guaranteed savings and services commenced April 1 '99 and as of June '01 there had been no unplanned production interruptions, Sarin reported. Enron leverages its purchasing power to reduce energy costs and manage their volatility, added Enron Vice-Chairman Marty Sunde.Collaborative supply chains
Supply-chain integration is being driven by the shift from competition between companies to competition between supply chains, said Bill Friend, retired vice-president of the J. R. Simplot Food Group, now principal of WR Friend & Associates (Boise, Idaho).The issue of "identity preservation" -- segregating genetically-modified commodities from their conventional types -- is also driving integration. In the past six months, Friend observed, 300 products were recalled after testing positive for Starlink, the corn variety genetically-engineered to resist the corn borer, "probably the largest single recall relating to a particular food issue." Government regulations concerning food safety, nutritional labeling, allergen labeling, electronic records and signatures, and trace/recall procedures are also driving integration. "Lot coding is evolving into smaller increments to accommodate recalls," he added.
Collaboration is the key to a competitive supply chain, Friend continued. But slow response by manufacturing and fragmented IT structures have constrained food companies from collaborating for e-business with suppliers and customers. To integrate with customers and suppliers, he recommended several strategies:
- Open data collection systems to supplier information.
- Import forecasting data from customers, and integrate it with advanced planning and scheduling software.
- Develop an export capability in planning system. This offers Available To Promise (ATP) information based on production capabilities. "Not many food companies are doing that," said Friend. It also allows automated scheduling for co-packers and suppliers.
- Develop an export capability in production systems, to consolidate product specifications and summarize production-run data.
Fast-response warehouse
Pat Lovesee, director of logistics and warehousing at Anchor Food Products (Appleton, Wis.) described the "pull" system implemented at Anchor when the company shifted from forecast-based production to market-driven manufacturing (FE Apr., '01). Anchor, a producer of frozen heat-and-serve appetizers, in 1999 installed a highly-customized Demand Management software package developed jointly with systems integrator On-Point Consulting (Hudson, Ohio). Demand Management allows the plant to operate as a flow process rather than a discrete process as originally mandated by Anchor's ERP system. The new system integrates production and scheduling functions; integrates touch-screen MMIs with a mainframe AS400 to execute production; and traces ingredients and finished products by lot. It also integrates plant-floor touchscreens via radio-frequency (RF) with MMI-equipped fork lifts "for quick response to production lines in minutes" with ingredients and packaging materials, and integrates via Web with suppliers for JIT deliveries. Web-enabled Advance Shipping Notices (ASNs) assure accurate supplier deliveries. Results, Lovesee reported: warehouse staff reduced from 128 to 84; weekly plant downtime reduced from more than 30 hours to less than 2 hours; 98 percent accuracy of inventory records; first-in/first-out control of ingredients and finished products; and accurate lot control.
