April 1999 Chilean Forestry Firm a 'Model' of Success By Jenny Summerour This is another in a series of articles based on interviews with recent Franz Edelman Award finalists. Geared toward practitioners, the articles strive to provide lessons the Edelman authors learned and some pitfalls they encountered during the course of their work. The full Edelman papers are published inInterfaces. Challenged by an expanding transportation system that was difficult to handle with its existing operations, Forestal Arauco, a Chilean forestry firm, decided to turn to a group of academics at the University of Chile in 1988 for some OR help. Jorge Serón and Pier Traverso, then head of production and transportation, respectively, at the firm, intitiated the effort. At the time, Forestal Arauco was producing 600,000 cubic meters of timber that was transported by about 90 trucks. Soon other Chilean forestry firms involved in the Forestry Production Group (GPF) joined the effort, led by Bosques Arauco. Andres Weintraub, Rafael Epstein and Ramiro Morales were asked to handle the project. "It was the first decision system developed to support operations for the forestry sector in Chile and was an important challenge to all involved," Epstein recalls. "When we looked at the problem the firms faced, we saw that the way they were handling it had deficiencies from several fronts," Weintraub explains. "From a managerial point, there were shortcomings in the way the firms organized some of the decision processes. In terms of data, there was a need to better organize its acquisition and handling; and, in particular, the decisions in all cases were very complex ones, so that the manual way in which the firms operated naturally led to lesser quality solutions." Developing the system It took Weintraub and Epstein about a year to develop the truck scheduling system based on a simulation model with heuristic rules, called ASICAM. "ASICAM had a dramatic impact on forest transportation," Epstein says. "A main reason for its success was that Jorge Serón and Pier Traverso became directly involved in the development and implementation of the system with a clear understanding of the problem and objectives." According to Epstein, this close collaboration between academics and managers was a common factor in the success of the systems that followed. The next system they developed was OPTICORT, a system that supports short-term harvesting decisions. One of the challenges in short-term harvesting is to match demand of products — defined by length and diameter — with the supply of standing timber. "The manual system used at the time by the firms led to a significant amount of degradation of timber, where higher diameter logs, suited for export or sawmills, ended up being used for pulp, with a considerable loss in value," Weintraub says. "A model seemed a logical way to get better schedules." "The system not only optimized the operational decisions in harvesting but also changed the way the managers looked at the problem," Epstein adds. "The model and its concepts became the natural language to discuss the operations. They had to negotiate the parameters, and the model would do the dirty work. The system had to run in a few minutes to allow discussion and negotiation; that was a crucial feature for the success of this tool," he says. The system, an advanced linear program, took about two years to develop. Again, the heavy involvement of Serón and Nieto at Arauco was a key factor of this success. The development of PLANEX and OPTIMED followed in 1991. PLANEX, which deals with the optimal location of harvesting equipment and access roads, required a sophisticated and interactive graphic user interface fed by a Geographic Information System. In addition, Weintraub and Epstein needed help from someone with knowledge in cable engineering for tower operations. They turned to John and Bren Sessions at Oregon State University. Once again, several managers and engineers from the companies were involved in the project, including Arauco senior managers Enrique Neto and Pedro Sapunar and senior managers from Bio-Bio and Mininco. OPTIMED, a tactical forest planning tool, was developed next. Lastly, MEDFOR was developed by Ramiro Morales as a long-range strategic planning model. According to Weintraub, "Using some form of mathematical model for each case not only provided better solutions, but also gave a strong impulse to acquiring the right data and creating a stronger management. "For example," he continues, "the system we developed for transportation required the creation of a centralized transport control center, which was much more logical to manage a fleet, and offered continuous data on available timber in all origins to feed the model. This need also reduced thefts and losses of timber." OR: operation roadblocks "At the beginning, many people did not believe that OR models could be useful for their decisions," Weintraub says. "We had to slowly create confidence in them." According to Weintraub, an important step was involving future users in the systems' development. However, in many cases, he lacked reliable information to feed the models. "In some cases we could easily create the data, such as travel times," he recalls. "In others, such as determining the volume of timber in a given area, there is a need to improve on inventory simulators the firms had." Despite the few problems they encountered, the University group for the most part claimed success. The systems reduced operating cost between 10 percent to 20 percent, according to Weintraub and Epstein. The systems also had a positive impact on the organization by forcing them to be more rigorous with the use of their resources. "It also decreased the time spent by professionals, who formerly were tied up doing manual calculations to obtain inferior solutions," Epstein says. "Now they are also more confident that they are making more correct decisions." Perhaps most important to the group was their influence in Chile. "This success has had a significant impact in the Chilean academia," Weintraub says. "More researchers have seen this field as an opportunity to add value to our export products." Lessons learned "The main lessons we learned were that we needed to look at the problems from the user's point of view — to see the projects as customer oriented," Epstein says. "What is their problem, what is the best solution for them? In practical terms, this meant involving very heavily the users in the development of each project." By including the users in their work, Epstein says he and his colleagues made sure that: a) The solution approach they were developing was comfortable for the user (i.e., the user had enough data, it was easy to use, the system run fit their decision process). b) They could provide answers that the user needed in a fast development, so the user could see relatively quick improvements. "This meant at first sacrificing sophistication of the mathematical models and quality of solutions," Weintraub says. "However, as the users saw the advantages of our approaches, they were willing to invest in further development to improve the solution process, including the mathematical models and the algorithms." c) The user was able to identify with the development of the project. d) The user could expect technical support at any time after they started using the systems. "We did not try to impose our agenda on the users," Epstein says. "On the other hand, we were careful to use OR models where we felt it could have an important impact. In this sense, we had a better feel for what could be gained from the use of the models, and in some cases we stopped some OR models the users thought we could work on but which we felt would not be worthwhile. "We always evaluated a project considering its impact and future improvements," Epstein continues. "We feel that the company managers respect our technical advice, considering them as independent opinion. This kind of long-term relationship is a great asset for us." The next effort for these OR practitioners is a joint venture with the Forest Research Institute (FRI) of New Zealand and John and Bren Sessions at Oregon State University to improve PLANEX and make it a standard in forest planning. Epstein says they expect to begin development in June. Jenny Summerour is managing editor of OR/MS Today. She can be reached via e-mail at jsummerour@lionhrtpub.com. Table of Contents OR/MS Today Home Page OR/MS Today copyright © 1999 by the Institute for Operations Research and the Management Sciences. All rights reserved. Lionheart Publishing, Inc. 506 Roswell Street, Suite 220, Marietta, GA 30060, USA Phone: 770-431-0867 | Fax: 770-432-6969 E-mail: lpi@lionhrtpub.com URL: http://www.lionhrtpub.com Web Site © Copyright 1999 by Lionheart Publishing, Inc. All rights reserved.