Tuesday, 27 September 2011

Managing an icehouse

When I was seven or eight years old, I discovered the word "icehouse". My grandparent's house was near a country estate which possessed an icehouse. According to my grandmother, there were very few icehouses in Britain. Somehow, the idea came to me that when I was grown up, I would write about the "very few icehouses". But when I grew up, I discovered that there are hundreds of icehouses in Britain, and many in other countries. I don't think that grandmother knew how many there really were.

Recently I came across a reference to icehouses which made me think of Operational Research. I have been reading the unusual book "A272: An ode to a road" by Pieter Boogaart. This is a book about places along the A272, things to see, information about the history, all written in a delightful style. And Pieter has a page about icehouses, and his frustration that he was not allowed to go into the icehouse at Petworth House.

The principle of managing an icehouse is simple, but there are conflicting objectives. The icehouse is filled with ice during the winter and the icehouse keeps the ice frozen all year long. During the year, food may be kept in the icehouse for preservation, or ice may be taken to chill food in the kitchen of the house. To work, the icehouse must not be opened too many times, else the ice will melt as warm air is allowed in. So the conflict; the cook would like to use the icehouse frequently; but by doing so, the aim of the icehouse will be overturned. So, I assume, most houses which used icehouses had rules (heuristics?) which suggested how often and when the icehouse might be used. Many of the owners of grand houses which possessed icehouses had second homes, or spent some of the time away, so access to the ice was only needed at certain times. It would be interesting to know if any such rules exist today.

Pieter Boogaart recounts the following story:
In Holland the efficacy of an icehouse was put to the test a few years ago. It was quite an exciting project, and a goodish number of people participated, each of whom was issued with a key. They were naturally proud of what they were doing and showed the icehouse outside and inside to friends and relatives. And they all came regularly to check on things. Which was exactly why it didn't work.
It would be interesting to find details of this project.

Monday, 26 September 2011

Water supply and the environment

Once upon a time, the citizens of a large city realised that their water supply was running low, and as the city was expanding, and the demand for water per capita was increasing, there was a problem of supply. So they looked to the mountains to the north and to the west, where there was high rainfall, and few people lived there (though there were many sheep) and asked some engineers to investigate where it might be possible to build a new reservoir.

And after some time, and after spending some of the city's money, they returned and announced that they had found an ideal site, which would mean displacing many sheep, but very few people. "What is more," they said, "although the site is about one hundred miles away, there is a convenient river which flows from the site towards the city, and could be used to convey the water part of the way from the reservoir to here. If you pay us some more money, we can devise a scheme to extract the water which will be released from the reservoir." And the citizens thought that this was a good idea, and the engineers devised a suitable scheme. Everyone was pleased because this solution was very good for the environment, as it did not need pipes to be laid through the beautiful mountains.

In due course, the people who lived by the river heard about the scheme, and went to the citizens of that city, and said: "We know that your reservoir is going to supply you with water. We would like you to help relieve us from the problem of floods. Please will you plan your reservoir so that it can hold back the water at times of floods?" They too were concerned for the environment, because they farmed and fished beside this pleasant river.

So the citizens listened, and as they wanted the people by the river to support their expensive water supply scheme, they agreed. But when the engineers who had devised the scheme heard of this, they were shocked, and said, "That means conflicting objectives!" For the benefit of the citizens, they explained. "If you want to supply water, the reservoir must be full. If you want the reservoir to hold back flood water, it must NOT be full. Water supply and flood protection are conflicting objectives."

"What shall we do?" wailed the citizens. "We need water. The people by river do not need floods." And one of the engineers announced, "The experts in conflicting objectives are called Operational Research Scientists. You must go and consult with some of them." So the citizens went with a bag of money to consult with a team of O.R. Scientists, who said that this was the sort of problem which they could help with, and appointed some of their number to work with the engineers.

The O.R. people, as is their habit, started asking questions about "the system", and as is their habit, started to read about other places in the world that such a conflict of objectives had been found. And they discovered that such a problem had been given a name, the "Noah-Joseph problem". Noah, in the Bible, had been involved with a flood. Joseph, also in the Bible, had been involved with a seven-year drought in Egypt. Floods occur very rapidly; droughts occur very slowly. They also discovered that there were some further environmental constraints and conflicts. First, the water from the reservoir would flow into a natural lake, and there was a town beside the lake which was liable to flood if the lake level was too high; therefore any plan must be constrained to avoid flooding this town (which incidentally had a very nice hotel where the O.R. people enjoyed their meals). Second, the new reservoir would have recreational facilities; sailing, fishing, and footpaths to permit visitors to enjoy the mountain scenery. All of these benefits to the environment depended on the reservoir being full. Third, although the people living by the river did not want to be flooded, they also did not want the river level to be "just below flood level" for lengthy periods, as this made the fields beside the river waterlogged.

So the O.R. people realised that there were two parts to the problem; there was a long term water supply problem, and a short-term flood mitigation problem. So they constructed two models. The first used records of rainfall, month by month, to feed into a model which determined a "Target level" for the reservoir for each month of the year, lower in the wet winter months, higher in the dry summer months. If the level was below the monthly target, then the control said "bring it up by reducing the releases", and if it was above, then the control said "let it down by increasing releases". The second model was concerned with controlling the river flow during times of flood risk, and modelled the control policy, hour by hour, through simulated run-off and releases from the reservoir.

And the staff employed by the citizens of that city were pleased with the work of the O.R. people, and used their plans for both long-term and short-term control of the reservoir.

In due course, the land had a severe drought, but the people of the city did not run short of water, thanks to the skills of the engineers and the O.R. people.

Some of this work is described in a more formal way in: D.G.Jamieson, D.K.Smith and J.C.Wilkinson "Evaluation of Short-term Operational Policies for a Multipurpose Reservoir System" Journal of Hydrology 28 p191-213 (1976)

Thursday, 22 September 2011

When were your neighbourhood streets designed?

An interesting, and counter-intuitive, correlation has been discovered by two urban geographers. The rate of fatalities in road accidents in cities in California depends on when the street grid was laid out. And, despite the belief that town planning has become better, the correlation is that older, more grid-like, layouts are safer. A possible reason is that grids make motorists stop and start more often, and so may drive more slowly. There's a popular account of this at:
http://www.theatlanticcities.com/design/2011/09/street-grids/124/

It emphasizes the O.R. principle of determining what system one is trying to optimize. Town planners may have designed curving roads, with tree structure and cul-de-sac roads to reduce through traffic but "the system" extends to the environment of work, shopping and leisure as well as where people live.

Tuesday, 13 September 2011

How times change

This has little to do with O.R., except to illustrate how a profession can change in half a century. What was O.R. like in the 1950s? There are people who can remember it then. The following extract comes from H V Morton's book "In Search of London" (1951) and is his edited account of meeting a man who had been an electrician in the 1890s and early 20th century. I found it fascinating.

He took a screw-driver from the pocket of his alpaca coat, he did something and, behold, the thing worked again !

"That was clever of you," I said. " You are evidently a wireless expert?"

"Oh no, sir, not an expert by any means," he replied. "But I understand electricity."

"Is it your business?" I asked.

"It was, sir, it was. Nearly sixty years ago, sir, I was apprenticed to a Bond Street electrician. In those days electricity was a dangerous novelty. The firm was a fashionable one, and it served the aristocracy; yes, sir, it served the highest in the land. It was a great thing to be employed by this firm. All the employees when they were sent out to mend a fuse, or to see a duchess about switching over from gas to electricity, had to wear frock coats and bowler hats. And we always had to wear gloves. Funny, isn't it, sir, to think of those days now, when an electrician will go into any house in Park Lane wearing blue overalls! But when I was a young man we were inspected by the head of the firm before we went out on a job. And we were never allowed to carry a tool-kit. They were carried by small boys. And we always had to go out in a four-wheeler."

"An electrician's life must have been a very gentlemanly one in those days."

"Ah yes, sir, that was the policy of the firm. As I say, sir, electricity was a novelty, and the aristocracy were very proud of their wonderful new electric light. But, looking back on it, how very primitive it was! I have often been ordered to sit until two in the morning beside the switch-board when a dance was being given in the West End, in case the lights fused. And one had to be properly dressed for that, too."

I was delighted with the little man. I had been right about him. He was interesting. His manner, his voice, his spurious gentility were all a reflection of the dead-and-gone Edwardian era, when even Bond Street electricians had to dress up and wear gloves when they served the aristocracy.

"And there was another rule of the firm," said the little man." If you went into a lady's bedroom and saw jewels lying about-and by God, sir, you did see some sparklers in
those days, for you know what the aristocracy were like coming home three sheets to the wind and flinging everything all over the place; well, it was a rule of the firm that you had to call the butler and ask him to see that the jewels were locked up. It wasn't that the firm distrusted its employees. Oh no, sir. As the head of the firm used to say to us, 'It isn't you we mistrust, but suppose her ladyship lost her pearls one night, she might blame it on the electrician!' "

And as we chugged down the Thames on that hot summer day the little man in the alpaca coat managed to build up for me a life-like picture of a dead world: "toffs" in silk hats handing out tips of golden sovereigns to young electricians, great ladies in low decolletage moving towards dressing-tables in a haze of champagne, butlers guiding the young man from Bond Street round the house, and the young man from Bond Street carefully drawing on a pair of silk gloves (provided by the firm) so that the delicate paint-work should not be damaged.

"It was a very different London, sir," said the little man

"Do you think it was a happier London?" I asked.

"Oh, undoubtedly, sir," he replied. "People were not spiteful in those days. And, say what you like, the aristocracy were the cream of the earth, sir. And in those days, sir, a sovereign went as far as ten pounds go to-day. But to-day, sir, nobody is happy."

And with these depressing words we arrived at Greenwich.

Cost benefit on the roads

">

The TED video uses a little cost-benefit analysis of the effect of stop signs where they are not really needed. An amusing bit of O.R. with a serious purpose!

Terminal 5 at London Heathrow

Tina and I went through Terminal 5 for the first time in August. We arrived in plenty of time (one of the downsides of living in Devon is the need to allow plenty of time for the connections to London airports). While we were waiting, having checked our hold baggage, we wandered from the terminal building back to the car-park and drop off points, which are in a separate multi-storey building. And there we discovered the personal rapid transport "pods" which carry people between the business car-park and the terminal. These pods could come out of a 1960s SF film; computer controlled vehicles, seating four people, running on a fixed track between the terminal and one of two stations in the car-park (not two car-parks, two stations in one car-park; the business users cannot be expected to walk too far, can they!) The journey took about 5 minutes.

It was the control rules which posed an interesting O.R. problem. At the terminal, there were four platforms where the pods could be boarded, and at the car-parks, there were two at each. So these constituted queues. Users arrive, and board a pod, if one is available. The pod departs, and -- if possible -- another one moves into place from a parking bay nearby. The control system moves empty pods around the tracks to wait in these three parking bays. So here are the questions:
(1) how many platforms should there be?
(2) what rules do you use to move empty pods around the tracks?
(3) how many pods do you need in the whole system?

We enjoyed the ride there and back again, and thought about the O.R. person who had thought through these questions.

Monday, 5 September 2011

Welcome to my new blog

For some time I have been blogging as iaoreditor.blogspot.com

As I am now retired from the position of IAOR editor, the time has come to launch a new, and more appropriate blog. So here we are.