Luk Van Wassenhove and IFORS Distinguished Lecturer

Luk Van Wassenhove was the IFORS Distinguished Lecturer at this year’s INFORMS conference. Here he is receiving congratulations from Tom Magnanti (right), President of IFORS.

Luk spoke on “Closed Loop Supply Chains: Past, Present, and Future”. Closed-loop supply chains are those where the supply chain bringing goods from consumers back to suppliers is also important. Luk gave an interesting historical perspective on this rather young field. He suggested that the field has gone through 5 phases:

1) Technical remanufacturing. Research into how to best remanufacture/resuse returned items, with little regard to how they come back or where they go after remanufacturing.

2) Valuing reverse logistics. Research and interest in how items coming back to a supplier can create value for that supplier. These models are more market driven than just waste stream recovery, and address the front end acquisition of items.

3) Coordinating decisions, bringing the forward supply chain together with the reverse supply chain. Once the magnitude of the problem is realized, the reverse chain impacts the forward chain, and vice versa.

4) Closing the loop, with dynamic decisions over the lifecycle of products. One aspect of this is the need to spend money to make money. Consider a “recycled” computer: one that is only a few weeks old is much more valuable than one that is months old. In such a case, investments might need to be made to increase the speed of the reverse supply chain.

5) The final phase of research, which perhaps should have been the first one, is “Is there a market”? While this area has increased in academic stature, and there are visible, but isolated, examples in practice, how can these insights be embedded in real firms. This brings in issues of accounting (how should returns be valued) and marketing (how should cannibalism be handled between original and remanufactured products)?

This was an ideal plenary session: broad, understandable and interesting.

Heading off to San Francisco …

for the INFORMS Conference (originally planned for New Orleans). US Airways cut back their direct flights from Pittsburgh, so it is through Minneapolis I go. I’ll try to post some of the interesting things I see at the conference (if I can get out of the bar long enough: see my comments on social capital below).

Thoughts on fun things to do in San Francisco? I think blues is on the agenda for Saturday evening.

The Community of Operations Research

For the past three or four years, I have given a talk at the INFORMS Doctoral Colloquium about the importance of social capital in a successful career. This talk is based on the book Bowling Alone by Putnam which argued that society (particularly US society) is becoming more detached, with fewer people engaged in the sort of interactions that lead to social capital. I think this issue is particularly important to OR people, since most of our best work comes by spanning boundaries, working with people in other fields in order to advance the other field while invigorating OR.

Some argue the web creates communities. I don’t think that is really true. The number of true, ongoing communities seems pretty small. One group I know if is a mailing list on an author I adore: Patrick O’Brian (the movie Master and Commander was based on his work). This is an active mailing list that has been around for a decade. Even for this “community”, the number of people who are active for more than a couple of years is astonishingly small, which suggests to me a lack of a true community. So is the technology too limiting or will face-to-face always matter?

Web Resources, then and now

Wandering around the web, I came across a site that had copied my web page on pointers in Operations Research from 1994, presumably to avoid the then-slow cross-Atlantic downloads. It is stunning to remember what life was like pre-web. That page has about 50 pointers, about half of which were “ftp” or “gopher” (a http/html precursor). And I think that page was pretty complete! Compare that to the current INFORMS Resources page which has about 1500 pointers (100-200 of which are broken, but I’m getting to those) and is by no means complete.

I’ve been thinking about the value of these resource collections in the age of google. I spend a fair amount of time with the Resources page, but even I, when looking for something in OR, start at google. I think Resource pages are useful to get a handle on an area, and I am glad we have one for OR, but it certainly isn’t the critical page it was pre-google.

Blogging

It’s been about a week that I have been keeping this blog, and so far I think it has gone well. It is nice to see the activity (I have opened up my sitemeter so you can check out the stats as well): there’s about 20 visitors a day, and people are clicking around to explore, so many are staying for some time. Not many comments, though: I must attract shy visitors! Comments, of course, are welcome: feel free to comment on an individual post or on the blog in general. Easy enough to do: just click on the number of comments associated with an entry, and you get a chance to both read the comments and to add one of your own.

Warranties and Inventory

Jay Swaminathan from University of North Carolina was visiting us today. He gave an interesting talk about how to set inventory levels when warranty replacement is a significant issue. This paper really hit a chord since I am working on my 5th(!) iPod. It seems without fail that my iPod fails after 3 or 4 months, requiring a return shipment and then a new iPod. The only good aspect of this is that the warrantee is reset, so an iPod originally bought in July 2004 is now warrenteed until November 2006. Who needs extended warrantees!

In any case, the paper (Jay together with Wei Huang and Vidhyadhar Kulkarni) made a couple good points: first, for fairly realistic data (they are working with a real, unnamed company), ignoring the warrantee needs in setting inventory can lead to pretty high stock-out charges. The second, less obvious but perhaps more important point, involved some new technology the company was planning to invest in. The company was going to put in a system whereby they would get detailed information on when an item was sold (and which item), rather than the aggregate sales values it was getting. This would give the firm an accurate distribution of the actual ages of the items in the field, rather than just the total numbers. It turns out that there was very little value in the more detailed distribution: aggregate information worked out almost as well. A good example of the value of analysis before making significant investments.

OR Exhibits

I just got back from Washington, where a friend of ours, a scientific illustrator, had an opening for his work at the AAAS. It is a very impressive show, with a mix of illustrations of insects, plants, dinosaurs, and extinct mammals. The attendees of the opening were a real mix. Some had a scientific background, but most were enthusiastic amateurs.

It is too bad that OR doesn’t lend itself to this sort of amateur interest. Martin Gardner‘s Mathematical Games column (and some of his successors) did appeal to the nonprofessional, and often had a strong OR flavor (that’s how I got my start). But having an exhibit, attracting a mix of people, seems unlikely.

The closest we come to this is some of the optimization art done by Robert Bosch of Oberlin (I am sure there are others, and I would like to hear about them). Bob works in dominoes, traveling salesman tours, and many other media to create art through optimization. See, in particular, his site dominoartwork.com for samples of what he does.

OK, that makes two things I want to do since I started this blog: solve world hunger through improved logistics and have an opening at a prestigious museum. Looks like this blog is going to cause more problems than it solves!

Gary Lorden at INFORMS


Most of the time, I can recognize the plenary speakers at INFORMS conferences, at least by name. The INFORMS San Francisco (nee New Orleans) conference has a guy named Gary Lorden speaking. Gary Lorden? Who the heck is he?

It turns out Gary is the mathematical advisor for the TV show Numb3rs, a crime show whose solution is generally based on mathematics of some sort. I found a review of one of his previous lectures: it sounds like it will be a blast! All the more reason to stay through Wednesday in San Fran.

OR in Africa

There have been a couple of things I have seen recently highlighting OR in Africa. The first was a presentation by Luk Van Wassenhove on humanitarian logistics (here is an article Luk, David Kaatrud and Ramina Samii wrote about the UN Logistics Centre). While not limited to Africa, that region is certainly a key area. Humanitarian logistics offers a tremendous mix of organizational and technical challenges, not the least of which is coping with the sheer number of governments, NGOs, and other organizations, all with different objectives and requirements.

An article in the October OR/MS Today by Jonathan P. Caulkins, Emily Eelman, Minoli Ratnatunga and David Schaarsmith talked about a conference in Africa, the Operations Research Practice in Africa conference. Held in Ouagadougou, Burkina Faso, this conference looked at opportunities for real-world problem solving using OR to address Africa’s issues. The summary brings up some interesting ethical issues (should OR be used to help repressive regimes?) and some area-specific OR issues (particularly the non-routine, non-replicable nature of crisis handling).

IFORS is holding its 2008 international conference in Africa: Sandton, South Africa to be specific. I’m traveling there in February to check out sites and speak with the local organizers. It turns out to be easier to get there than I thought, with direct (one stop for fueling) flights from JFK, Dulles, and Atlanta. I am also going to spend a few days in Capetown.

The issues brought up in these papers do make me wonder if working at scheduling professional sports leagues (my own current research) is the most valuable thing I can do with my life!

Robust Optimization

Dimitris Bertsimas gave a talk here today on robust optimization. One question he asked was (paraphrasing) “What do you do when reality refuses to match up to the model?”, which I think is a great question. So much of what we do seems to be fragile (think cascading effects of a snowstorm in Chicago stranding travelers in Miami) when we know that are models are based on data that is only an approximation to reality. Robust optimization (roughly, optimizing with an assumption that no more than a certain number of data points are wrong, and each is wrong by no more than a fixed amount) is one way to attack this. Stochastic optimization is another. I am not sure we have found the right method yet (though Dmitris’ work is extremely impressive!).