There was recently a post in SIMSOC (Newsgroup discussing the use of Computer Simulation in the Social Sciences) asking for examples of agent-based models that have actually been used to support policy decision making or for other purposes "in the real world". The poster (Nigel Gilbert) was looking for examples which have been used for practical purposes rather than those that have been designed to be applied but have never in fact been used.
The post and a summary of responses can be found here:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=simsoc;1cf16ad3.1312
Are there any such projects related to OR that you have worked on or know of? Mine are unfortunately all examples that have been designed to be applied but have never in fact been used. Is this a general tendency?
I just remembered that during the "ESSA Summer School 2013" Abhijit Sengupta gave a lecture on "Social Simulation in Industry". He provided several examples (during the lecture and in a private discussion afterwards) where ABMs have been used for real world decision making in industry (e.g. Fruit Juice Market in Switzerland).
Presentation slides are available here:
http://www.tuhh.de/t3resources/essa/Abhijit_Sengupta.pdf
There is a large project at Aarhus University where ABMs will be used for evaluating how the construction of off-shore wind farms affects harbour porpoise populations. Scenarios will be run corresponding to wind farm construction in different parts of the North Sea (the DEPONS project, see http://pure.au.dk/portal/da/projects/effekter-af-forstyrrelse-paa-marsvinepopulationen-i-nordsoeen-depons(438d63df-7d73-469b-9306-f3c293bd64fc).html).
The ABM used DEPONS extends the one presented by Nabe-Nielsen, J., Sibly, R. M., Tougaard, J., Teilmann, J., & Sveegaard, S. (2014). Effects of noise and by-catch on a Danish harbour porpoise population. Ecological Modelling, 272, 242–251. doi:http://dx.doi.org/10.1016/j.ecolmodel.2013.09.025
There are of course the pedestrian and vehicle traffic simulation projects from Dirk Helbing and Stefan Lämmer in Dresden and Zurich, which have often actually been applied. A good part of this work is listed on the wikipedia page of Helbing: http://en.wikipedia.org/wiki/Dirk_Helbing
I actually think there are many cases in which research projects led to actual decisions... I only provide a few examples in an area I know well (pedestrian simulation).
One of these cases, about the management of a public event generating potentially problematic crowding conditions, is described in a paper by Micheal Batty (Michael Batty, Jake Desyllas, Elspeth Duxbury: The discrete dynamics of small-scale spatial events: agent-based models of mobility in carnivals and street parades. International Journal of Geographical Information Science 17(7): 673-697 (2003) - http://dx.doi.org/10.1080/1365881031000135474).
Fewer cases led to the birth of spin-off companies that turned out to be successful in the market (a nice example is represented by TraffGo GmbH - http://www.traffgo-ht.com/en/company/index.html).
Sometimes models developed in the academia have been later encapsulated in the engine of commercial products (e.g. the Social Force model of the above mentioned Prof. Dirk Helbing in the VisWalk pedestrian simulator - http://vision-traffic.ptvgroup.com/fileadmin/files_ptvvision/Downloads_N/0_General/2_Products/3_PTV_Viswalk/EN-UK_PTV_Viswalk_Brochure.pdf).
Similar examples can be surely found in the traffic area, and I'm pretty sure in other fields of application.
Sincerely
Giuseppe
My first ever OR project, soon after I joined the UK department of the Environment and Transport OR group in 1976, was a national cost-benefit analysis of snow and ice clearance operations on UK motorways. I was the only one working full-time on the project, but worked in a team with two more experienced OR colleagues and an administrator. They ensured that all 14 of our working papers were reviewed at monthly meetings of the steering committee which comprised our team together with the director and several other members of the WInter Maintenance Directorate.
At the core of the project was a simulation model I developed predicting hourly snow depths in each lane, delays and increased accident risk for any given pattern of snowfall, traffic flow level, numbers of snow ploughs available and snow clearance policy.
Our recommendations which were accepted by the client and implemented included halving the amount of salt spread (as long as the salt was down in time additional amounts had very limited effect). This meant that they could reduce the number of motorway depots and create much needed space for additional service stations.
We recommended different resource levels in different areas of the country depending on weather conditions and traffic flows, but in general recommended purchasing new faster vehicles but that overall fewer vehicles were needed. Overall, these recommendations reduced accidents and delays, but also saved the service over a million pounds (in 1976 money!).
My view is that the decision to have regular meetings with the steering group to check our assumptions and predictions were reasonable was critical to the successful implementation of this OR study.
Max Moullin, Director, Public Sector Scorecard Research Centre, Sheffield.
www.publicsectorscorecard.co.uk
Max is a Fellow of the Operational Research Society and the Chartered Quality Institute.
OptTek Systems (www.opttek.com) offers OptForce, a tool for strategic workforce forecasting and optimization. OptForce embeds an agent-based simulation of an enterprise's workforce and effects upon it from demographics, employment markets and employer workforce management actions (e.g., training investments, retention incentives). A number of commercial and government organizations use OptForce.
http://www.opttek.com
Many thanks for all your responses. It is good to see that ABM/S is in fact used in OR to support real world decision making.
I am currently editing a JASSS Special Section on "Engineering Agent-Based Social Simulations". As this is a related topic and some of you might also be active in "Social Simulation" I thought I should mention it here. The submission deadline is the 31 March 2014.
CfP: http://www.cs.nott.ac.uk/~pos/docs/pos-CfP-JASSS_EngineeringABSS.pdf
Of course, if you are prepared to accept that some of the combat simulation are infact agent-based models (or a variation of); then there are plenty of examples of simulation being used for (very expense) pronouncement decision making. You will need a contact in MOD(Dstl) or DOD (or equivalent elsewhere) to see what is unclassified though i believe there are examples in JDMS and MORS.
I know a friend has applied ABS with DES on an actual radiology center.
https://www.researchgate.net/profile/Mohammed_Abdelghany
- Badges
- Science topic
- Similar topics
- Academic Journals