Friday, 14 June 2013

Announcing simulation model "Conference optiloo"

Conference optiloo will be a visual simulation model designed for large conference centres where the management wish to reduce congestion at their toilets at the end of addresses and speeches.  It should be based on the premise that most conference centres have used national or international guidelines for the number of washrooms on the assumption of a 50:50 split of genders among the delegates.  These guidelines should be adequate for a wide range of splits of genders from roughly 20:80 to 80:20.  Conference optiloo will be designed to assist the planning where the gender split is more heavily weighted than these figures suggest.  Examples might be conventions for women in management or organisations where males are most numerous.

The simulation model will use observed data from conferences to measure the demand for washrooms after long plenary sessions, giving typical time-dependent rates of arrival for users during the breaks between such session.  Using publicly available data for the gender-specific distribution of occupancy time of washrooms, adjusted if empirical evidence supports the hypothesis that these times are reduced when the user is aware of a queue outside, the model will provide the conference management staff with visual information about the build-up of queues at restrooms.  The model will offer such staff the ability to change the gender allocation of each washroom area from male, to unisex, to female, again making allowance for empirical behaviour of patrons in a unisex facility. A fast-running model in the background will complement the visual output with statistics from several independent runs of the model, which can be used to make informed comparisons of alternative systems.  Its use could be extended to large theatres, sports grounds and auditoria

This simulation model is bound to have widespread application as seen in this blog entry.  Please note that the model does not actually exist, nor that I am not planning to research or write it - but the idea is offered to anyone who wishes to develop it further - but please credit me with the idea!

Thursday, 13 June 2013

Silicon Valley Men's Room Problem

Sooner or later, most lecture courses on queues mention the problem of queues for toilets.  Generally the focus is on the provision of facilities in public buildings, and the difference between the genders.  One writer summed up the problem that exists in theatres and similar places where the demand for toilets peaks during a short period of time; if there is a queue for the men's room, as well as for the ladies' room, then there is a real design problem.

The UK national newspaper, The Independent, published a picture and article yesterday with a twist to this story.  It showed Dan Ackerman's picture (above) of a queue for the men's room, and no (visible) queue for the women's.  It was taken at the Worldwide Developers Conference (WDC) in San Francisco.  What does it say?  The majority of delegates at the conference were male, but the provision in the conference centre assumed a more even mix of delegates.

The picture started a Twitter account whose aim is to see enough women in software development that there would be a queue at the women's restroom.

But OR is not just about modelling queues; it should aim to reduce congestion.  What might be done in the conference centre for such an event as the WDC, where the organisers know that there will be an imbalance?  The same problem applies in reverse for conferences targeting women.  If there really only two restrooms, then there is little possible unless all the provision is made private (i.e. no public urinals).  But at a large conference, then one could designate some restrooms of one gender to be temporarily for the other.  Yes, the minority might have to walk, but the majority would not have to cross their legs in the queue outside the doors.

Step forward someone who attended the WDC and can give figures for the number of accessible restrooms for each gender, and step forward someone who could model the system.