February 5, 2004
Like any successful IT project, implementing optimization software that's going to benefit the business requires more than uploading software, inputting a bunch of numbers, and running the program. Operations research expert Michael Trick, professor of operations research at Carnegie Mellon University, offers these tips:
• A successful optimization project requires a deep understanding of the algorithms involved and what they're capable of, as well as a thorough understanding of the organization's business processes and goals. • Make sure the data put into the program is accurate and up to date. Trick says companies often make the mistake of assuming data files--such as key financial information--are timely and available, only to find out they're not. "Even before the model is created, companies need to make sure the data is there and is correct," he says. • Don't try to tackle every problem at once. Identify the important problems and critical processes, and focus on those first. And take the time to revisit those critical problems you couldn't solve last year, as well. "Operations research is a very quickly developing field. What didn't work a year ago might work now, so you can always go back to problems that couldn't be solved before," Trick says. • Find a champion for the project. United Parcel Service's success with operations research and optimization software lies in the fact that they're supported at the highest levels of the company. In fact, the company's current operations research group was founded by chairman and CEO Michael Eskew in the mid-1980s.
About the Author(s)
You May Also Like