Optimizing Production Planning: Lessons from a Small Agribusiness

Friday, May 19, 2017

The Jan de Wit Company, www.jandewit.com.br, is a wholesale producer of bulb flowers in Brazil. In 2000, when the company began to consider optimization to help its production planning, it had 18,745 square meters of greenhouses, 1,032 square meters of cold-storage rooms, a team of approximately 30 employees and US$80 million in annual sales.

It is difficult to overstate the complexities of production planning at Jan de Wit. As with many businesses, planning begins with an attempt to project demand. For Jan de Wit, this involves two types of sales: auction and intermediation. Auction represents a daily cash market at which distributors can bid on flower purchases. It accounts for approximately half of Jan de Wit’s sales. Quantities and prices can vacillate considerably, reflecting such things as fashion trends, economic conditions, and competitors’ production volumes. Prices, of course, fluctuate, affecting Jan de Wit’s optimal product mix. 

The other half of Jan de Wit’s sales are through intermediation, which operates like a futures market. Specialized agents negotiate buy-and-sell contracts between distributors and producers for the short, intermediate, and long term. Based on history—and what Jan de Wit can surmise about the market—the company attempts to identify market opportunities by analyzing weekly sales quantities and prices for each lily variety.

To produce an optimal production plan, Jan de Wit Company must schedule the correct planning of the right bulbs during the right week in the right greenhouse environment to meet project market demand, which must include seasonal fluctuations for such holidays as Easter, Mother’s Day, and Christmas. Some additional constraints on planning decisions include bulb inventory and production-cycle variations stemming from plant variety, bulb size, bulb origin, sprout length, and planting week. Technical requirements produce yet one more layer of complexity, with such constraints as number of bulbs per pot or box, bulb spacing, and bed-usage limitations associated with each type of greenhouse.

The way in which optimization first found its way into the Jan de Wit Company illustrates how operations research concepts are spreading to smaller companies. José Vicente Caixeta-Filho, a professor at the University of Sao Paulo, published a short article in a student journal entitled, “Modeling, Through Operations Research, in Agribusiness.” Jan Maarten van Swaay-Neto, a flower-business management consultant who had not previously heard of operations research, read the article and called Caixeta-Filho to discuss possible applications to the flower business. He went on to take one of Caixeta-Filho’s classes on linear programming, which was being offered to graduate students in applied economics. At the end of the course, he wrote a final paper entitled, “Gladiolus Bulb Production” and extended an invitation to Caixeta-Filho to develop more accurate approaches to applying mathematical models to flower production problems.

This initial collaboration led to further efforts to apply the techniques, first at a company named Terra Viva and then at Jan de Wit. Reflecting today’s abundance of computing power, the decision-support system was programmed on a Windows-compatible computer, using Visual Basic and Microsoft’s Access database. The team used linear programming to maximize contribution margin (revenue minus variable costs). The actual set of equations in the linear program generated a solution matrix involving 120,000 rows and 420,000 columns.

The results of using optimization software to aid production planning speak for themselves. The first year that Jan de Wit used the software, revenue grew 26 percent, while contribution margin improved 32 percent. Return on owner’s equity increased from 15.1 percent to 22.5 percent—a 49 percent improvement. This growth was managed with the addition of a single employee. All this was accomplished in spite of the fact that during that year the Brazilian flower market experienced excess capacity and reduced auction prices. The company also found that by improving its planning and control of production it was able to increase its short-, medium- and long-term supply agreements, locking in prices and profits years in advance.

Finally, there was the saving in planning time. Prior to implementing the optimization software, Johannes de Wit, the general manager and owner of Jan de Wit Company, planned production himself. With the new system, he found that he was able to delegate the planning process, which now takes only a fraction of the time it once required. Clearly, implementation of the planning system has increased the company’s competitiveness. De Wit commented, “Companies in the flower business that don’t wake up to planning-and-control systems risk almost unsurpassable capital losses, endangering their continuity and damaging the market.”

This post is excerpted from The Optimization Edge: Reinventing Decision Making to Maximize All Your Company's Assets (McGraw Hill) by Steve Sashihara, www.optimizationedge.com.

Photos taken from Jan De Wit company video at www.jandewit.com.br