CO 370CM 443 Deterministic Operations Research Models Henry

CO 370/CM 443 Deterministic Operations Research Models Henry Wolkowicz hwolkowicz@uwaterloo. ca Office: MC 6065

Today’s lecture • What is Operations Research (OR)? – Examples and success stories • What you will learn in this course – Course outline • Administrivia • A little LP review + LP formulations

What is Operations Research? • Enterprises typically consist of various units whose operations need to be coordinated – A typical problem: how to distribute resources across units so as to maximize efficiency • OR is the discipline of applying analytical tools based on mathematical models to help take better decisions in managing these operations (the “science of better”) • Typical decision-making process in OR entails: 1. 2. 3. 4. Gathering available data Building an abstract mathematical model Solving the mathematical problem Supplying the results to management Feedback: may need to enhance data or model

A Brief History of OR World War II led to the birth of OR • Scientists and engineers used mathematical tools to Perhaps the single deploy, manage, and analyze military operations, e. g. , most important catalyst in the – deployment of the radar advancement of OR – management of convoy and submarine operations • Many important developments took place in this period. Most notably George Dantzig invented the simplex method for solving linear programs (LPs) in 1947 – One of the first applications was the diet problem: given foods with varying nutrient amounts, plan a diet that satisfies the desired nutrient requirements + the food-amount constraints, and minimizes cost

Brief History (contd. ) After the War: • • • Industrial boom led to rapid increase in size of corporations – growing need for systematic decisionmaking tools Managers began to realize both, the modeling power of LPs and OR tools, and their potency in improving efficiency even under existing technology; applications increased manifold Serendipitously, great advances were made in computational technology, which allowed one to solve problems of ever-increasing size via OR tools

Where is OR Today? • • Immense computing power available readily and fairly cheaply, e. g. , PCs A half-century of research in OR has led to: – very good theoretical understanding – various software packages, e. g. , CPLEX, LINDO, XPRESS-MP, being available that can be used “off-the shelf” OR is everywhere – from booking an airline ticket, to checking into your hotel By the end of this course, you will be able to solve real-world problems on your PC using these OR tools!

OR in action: Optimizing Harbor Operations Picture of container terminal Altenwerder in Hamburg One of the most modern terminals: handles ~ 2. 4 million containers yearly All internal traffic and storage cranes are automated.

Harbor Operations (contd. ) • • Overall objective: minimize vessel waiting time Containers are transported between vessels and storage area via automated guided vehicles (AGVs) Moehring et al. 2004 give a fast algorithm based on shortest-path computations that works much better than previous adhoc methods Subproblem: compute AGV routes that are free of conflicts and jams while maximizing container throughput. Also routes have to be computed in real-time.

Some other success stories • • • Continental Airlines saves ~ $40 million by using OR tools to near-optimally reassign crews after disruptions Texas Children’s Hospital used nonlinear optimization to monitor healthcare contract negotiations Athens Olympic Organizing Committee used logistics, optimization tools to manage its resources and plan the 2004 Olympics; estimated savings: $70 million Philips Semiconductors saves ~ $5 million by using stochastic multiperiod inventory theory to handle demand uncertainty Many more applications on “The Science of Better” website on Syllabus page

Yet, it is always surprising (to me) that there are many applications out there that are tailor-made for the use of OR tools that are still tackled by adhoc methods. Article in Science. Daily titled “Techniques for making Barbie Dolls can Improve Health Care” talks about how OR tools will soon come to be adopted to improve health care delivery Researchers working on optimizing Air New Zealand’s crew scheduling reported that “many airlines still use heuristic or manual methods”

What you will learn • Mathematical Modeling – learn a variety of ways of modeling realworld problems as structured mathematical problems • Solution Methods – learn to use powerful optimization tools to solve the problems arising in your mathematical models

Course Outline • Linear Optimization Models – Formulations, Sensitivity analysis • Stochastic and Robust Optimization – Decision-making under uncertainty • Integer Optimization – Formulations, Solution methods • Network models and algorithms – Max-flow, min-cut, shortest paths • Dynamic Progr. ; And Nonlinear Progr.

Administrivia • Course webpage: www. student. math. uwaterloo. ca/~co 370 – Check here for announcements, assignments, lecture notes, important dates, … – Email: co 370@student. math. uwaterloo. ca • Reading material: CO 370 course notes; highly • recommended Software: AMPL; trial version available from course homepage; full version on student. math UNIX machines

• Grading Administrivia (contd. ) – Assignments: 15% – Project: 20% – Midterm: 20%, 6 (+1) assignments Final: 45%

Administrivia (contd. ) • Cheating policy – students caught cheating get no credit for that assignment/project; will be reported to assoc. dean – you may discuss the assignment with others, but must write up your own solutions • Feedback, comments, suggestions, questions are always most welcome. – Come see me at MC 6065 – Office hours: Wed. 2 -3 pm, Thurs. 3: 00 -4: 00 pm – Office hours of TAs on course-website

LP review: Definitions Linear programming problem: – problem of maximizing or minimizing a linear function of a finite number of variables – subject to a finite number of linear constraints: £, ³ or = constraints max/min f(x) = c 1 x 1 + c 2 x 2 + … + cnxn £ subject to ai 1 x 1 + ai 2 x 2 + … + ainxn ³ bi "i=1, …, m = Feasible point: xÎRn s. t. x satisfies all constraints Feasible region: set of all feasible points P = {xÎRn: x satisfies all constraints}

LP review: more definitions Decision variables: Objective function max/min f(x) = c 1 x 1 + c 2 x 2 + … + cnxn subject to a 11 x 1 + a 12 x 2 + … + a 1 nxn £ b 1 a 21 x 1 + a 22 x 2 + … + a 2 nxn ³ b 2 a 31 x 1 + a 32 x 2 + … + a 3 nxn = b 3 should completely describe all decisions to be made Constraints Optimal solution: feasible solution with best (max/min) objective-function value Optimal value: objective-f’n. value at an optimal solution