S/he triumphs who knows when to fight and when not to fight: using combinatorial optimization and data science in heavy industrial settings
Thursday, November 12, 2015
11:45 am - 12:45 pm
Gross Hall 330
Elliot Wolf (Lineage Logistics)
Mathematical optimization provides exact solutions to problems in heavy industry of massive economic importance, from logistics network planning , to asset and labor scheduling, to minimum energy/minimum error industrial control to space utilization optimization. Despite heavy industry¿s problems being textbook applications of mathematical optimization, it seldom relies on it. And when it does, it is often only on the margins of a process or problem. This talk examines mathematical optimization in practice, via a case study where bin-packing was used to design a large industrial facility to maximize storage capacity. In addition to an overview of the math and data science involved, the talk explores compromises between competing objectives (many of which cannot be expressed mathematically), using tractable relaxations of intractable problems, difficulties defining constraints, and acting amidst uncertainty. All are necessary to getting the job done, in addition to sophisticated mathematics. About Lineage Logistics: Lineage Logistics is the second-largest cold storage warehousing network in the world, with 111 (enormous) warehouses in 22 states. The company stores temperature-sensitive materials, blast freezes fresh products, brokers temperature-controlled transportation, and provides a variety of value-added logistics services. Lineage owns approximately 25 percent of the third-party cold storage capacity in the US, with annual throughput in excess of 15 billion pounds.