Stories tagged Applied Operations Research

Jan
26
2010

When we decide to go from point A to point B, we have a plan. If driving, we would decide which route to take, which intersections to avoid, and we will estimate how much time it will take. As we do drive from A to B, we will, intuitively and deliberately, make changes – we might change our route because of a traffic jam, or speed up to try and cover time lost. Thus, as we encounter new information and unexpected situations, we, intuitively, make changes to our plans, for the better.

Scientists and engineers have been trying for years to bring this capability – of automatically changing plans – to machines. Research in robotics has focused on equipping machines with sensors and radars to detect changes and respond to that change. For example, Toyota Motor corporation equips its Lexus LS460 with a radar which detects vehicles and obstacles on the road ahead. When the radar detects the possibility of a collision, the vehicle retracts the seatbelts, warns the driver, and applies brakes to reduce collision speed. Such systems are dubbed “sense-response” systems, and today are even available in automated vacuum cleaners such as the Roomba. However, taking the sense-response system a step further, such that it is applicable to entire supply chain and production systems has proven difficult.

A key breakthrough in developing automated “sense-response” systems, that would enable developing automated production decision support, sales and marketing recommendations, and several other automated systems has recently been achieved. Dr. Nazrul Shaikh, a post doctoral researcher at University of Pennsylvania has worked on the characterization of the planning and re-planning problem and have developed solutions for several classes of sense-response problems. The research, which has survived the extensive peer review from academicians, in now all set to make an impact to homeland security and corporate America. It will change the way the automated planning systems are used and implemented in industry.

Dr. Shaikh’s quest to characterize the sense-response problems began in 2002 when they were devised an approach to develop an intelligent sense response system for supply chain event management for IBM. This was in the wake of 9/11 when Corporate America realized that they need to be able to handle exceptions to plans better. The research, called Exception Analytics looked at the deviation between the planned and the actual values of several system variables and attempted to infer the “action” required to compensate for the deviation. Applications have been developed for supply chain event management, homeland security, quality control and marketing planning. Keen interest has been displayed by several companies, especially in marketing where the consumer and competitor dynamics influence the demand drastically. The basic concepts have been published in several journals after clearing rigorous peer review.