A friend of mine, Michelle Aden, recently asked me to write about some aspect of Open Source for a conference on that topic. It occurred to me that research had recently enjoyed becoming part of the Open Source community, so I looked into writing about that.
Here is the resulting paper. It’s also available in PDF and Word format.
I should note that considerable editorial help came from Sherrie Bolin, making it a much cleaner read!
This section has talks and papers I’ve given. The sidebar has pointers to several of these, and we’ll begin filling out this section with more detailed discussions of the papers and events leading to them.
The Self-* workshop was held the end of May in Bertinoro, Italy. It was held in an absolutely fabulous facility owned and operated by the University of Bologna in this gorgeous hilltown.
The purpose of the workshop was to look at a wide range of self-xxx ideas: self organization, management, healing, deploying etc.
My paper was on the BeerGame investigation done earlier with the Santa Fe Institute’s business network.
I publish articles at the O’Reilly site now and again. In the future, I’ll write more detailed entries here about them, but for now, here’s a list of the past ones:
On the Wave/Particle Duality of Random Events
Java Scripting: Half the Size, Half the Time
NetLogo 2.0: Graphs, Nodes, and Edges
Harmony Internet Remote
Java Geography: OpenMap
Beyond Crypto: Adaptive Security
Domain Maddness!
NetLogo: Experimental Worlds in Silico
Math: The Next Big Thing
Hard to believe, but one of the classics of supply chain research is a board game! What’s more, its about beer! It is a simulation of a beer factory and its supply chain, all the way to the retailer and customer.
The interest in supply chains comes from the observation of the extreme volatility they exhibit. This has come to be called the Bullwhip Effect.
In a nutshell, very surprising swings in customer demand appear in even the most predictible products, Pampers, for example. You’d think B babies, P poops/day, U utility function (sometimes you resuse them so U=1.25, say) of the pamper would yeald a rock stedy daily demand of B*P/U. Nope. Wild swings in demand instead.
Our research consisted of reproducing the original Beergame in RePast, a java based simulation environment, and then tweaking it to look at visibility up and down the chain, and the impact of mesh chains rather than just a simple chain.
Some sources:
While at Sun, I began investigating the use of graph theory for peer networking. This was a great couple of years with Duncan Watts delving into “small worlds” and then Bernardo Huberman’s team working out local knowledge search within power law networks.
This resulted in a several week long validation of the power law search algorithms, using a large number of network topologies.
Here are some pointers to a nifty project we started at Sun. Local Knowledge computing and networking looks into techniques for managing them without central control. This s the heart of peer-to-peer systems.
The project begain with a survey of the field which looked at motivation, history and potential deliverables. We wrapped this up into two interesting slide presentations.
My first serious introduction to complexity science was the Santa Fe Institute’s summer school of 2000.
Sun is a member of the SFI business network. This includes a place at the Complex Systems Summer School, a graduate level multi-discipline gathering of mainly students, but also including business network and other affiliates to SFI.
I’ve got several pages and talks relating to the experience:
Powered by WordPress