End of The Road

The MBA Second Life exercise ended. Ended up a close run between two groups, both making sensible decisions. The third group slowly slipped behind, making some strange decisions, so not sure how educational this has been. Two groups getting marks of 84 and one getting 27: no tendency for group marks to regress towards the mean here! It will be interesting to read the students' reflections on the exercise. The exercise was badly affected by the difficulties of getting the Second Life viewer working on the Business School's pcs now that they have been upgraded to Windows 7. Just ran in to Frank Lassard who suggested that I should have used the Phoenix viewer instead as education had similar problems, so hopefully the problem will be sorted when the exercise is used for an honours course after Christmas.

Impact of Assessment

This is the first time that we have had the performance of groups in the simulation tied directly to a mark that the students get for their operations management course. The logic for this was that by making it count a small part towards the course mark students would be motivated to analyse the decisions they make. It was decided to link it to the amount of cash their company has at a cut-off date next week because tying it to assets leads to end game strategies of running up high inventories. The students also get a good brain-washing on their courses about Lean. The effect seems to have been a tendency to cut inventories and maximise cash in the short-term, but then find that this constrains a policy to increase throughput by cutting prices because there is not the material coming down the pipeline to allow it. Simple insights like realising that with no uncertainty about the delivery of raw materials it makes sense to make the frging buffes stocks either zero or a multiple of the current batch quantity: if the batch quantity is six a buffer of two does not allow a batch to be brought forward in response to an unforeseen increase in sales.

Tying the results to assessment has also limited the role of the lecturer in coaching the groups on what they are doing. Where in the non-assessed past the lecturer would email students to inquire why they made decisions and gently pointing out alternatives, the lecturer now feels that any coaching would give the coached an unfair advantage. He has put down a series of questions on the website though to ry to get them to think about what they are doing, but I cannot see evidence that they have read any of it.

Excel Spreadsheets

Asked this morning to set up the decision screens in the factories to produce Excel spreadsheet versions of the historical data to help the students "reflect" on how the exercise has gone. (I think "reflect" might be a euphemism for identifying the guilty...). Compared to some of the weird requests I am asked to make, this was a piece of cake. Just used ExcelLibrary (http://code.google.com/p/excellibrary/) to dump database tables into Excel spreadsheet files on the server then put links into the ASP.NET website. The files are enormous, so that should keep the students occupied for an hour or two.

Prima Colazione del Cane

Spotted today that one of the MBA students had cut a price yesterday for one item to one customer to £3000, when I presume they meant to cut it to £30000. As this is below the scrap value, result was massive surge in orders from this one customer for this one assembly. Pointed this out and was told this rapid descent into mayhem would be "educational" for the students. I then pointed out that having large numbers of orders placed on this item everyday feeding through to the requirement for forgings would generate inventory at a rate that would crash the simulation in eight hours, at which point the educational attractiveness of chaos went out of the window. The boss dived in and reset the value and I had to write a mod fast to wipe out the orders generated by the assumed mistake. I am not sure how pedagocically sound this action is.

The planning system has still thrown a surge of inventory on one family of parts into this factory, so maybe one of the students might spot this and realise increasing the capacity, if only short-term, might stop it all gumming up.

Second Life MBA Class

I have been drafted in to sort out some odd little technical quirks with the simulation in Second Life. It is now being used by the current MBA class. Nice to see that the edicts that students should not appear as non-humans are being ignored. We seem to have a rich collection of furries... plus a robot, who I think is Japanese, so no cultural stereotypes here!

The simulation is set up with high inventory and low resource utilisations, just to make sure it does not spiral out of control while students are getting acquainted with it. But this means it takes time for sales managers to realise that cutting prices a bit will start to increase sales and thus get the factories running more efficiently.

Opensim

Following the price hike from Linden I have been asked to look at moving our virtual worlds presence from Second Life to OpenSim. It will be cheaper and demonstrate our touching faith in open software. On the suggestion that we close down the SL region immediately and shift everything across for the start of the coming semester our faith in open source in general and OpenSim in particular was found to be less rock-solid, so the plan is to run them in parallel in the coming year. Aim is to run OpenSim on a virtual server with 4GB RAM and Windows 2008 R2, but not sure exactly how much virtual real estate this will give us, and get it set up by the start of semester in mid-September. What could possibly go wrong?