Transcript Sept 2008
Gareth Mitchell: This is the official podcast of Imperial College London. And welcome along to our September edition. I’m Gareth Mitchell. And this month, the best shot at predicting the worst case scenario with the computer simulations that tell you what to expect when the unexpected happens. That’s in just a moment. Also in this edition, the wealth of networks and a wealth of ideas as Imperial hosts a meeting on the future of the internet.
John Darlington: The internet has had dramatic effects on people’s lives and productivity of industry and the way government is run but there’s a feeling that we’re due for another revolution in the development of broadband, the convergence of mobiles, with the internet in the way we do things.
GM: And not so much scalpel please but poetry please. And Imperial student offers a poetic treatment of a surgical theme. Specifically the human side of the science of organ donation.
Georgia Gale Grant: I received a gift back in 72, it came from someone I never knew.
Nadey Hakim: I always say the kidney is priceless.
GGG: It came with no card, ribbon nor bows but sent to me, that’s all I know.
NH: You can’t create it yourself. The dialysis machine is an artificial kidney.
GGG: It didn’t arrive in a box or a sack and is a gift one would never send back.
GM: And as ever, we’ll have some headlines from around the College for you. That’s all to come right here on the official podcast of Imperial College London.
GM: Right then, you run an airport, a major manufacturing plant or possibly a hospital and suddenly there’s an emergency like power failure of the telecoms network goes down. People’s livelihoods or maybe even their lives depend on your actions. So what should you do? Well, I must admit I don’t really know. The trouble is neither do the people who actually are in those crucial positions of responsibility. And who can blame them? After all it’s not very good for your career development to switch the lights off for an hour and see what happens. Whilst serious disruption to infrastructure is thankfully rare catastrophic failure is theoretically possible. So it’s a good thing that it’s theoretically possible to predict the previously unpredictable with the help of some powerful computer models.
So Imperial is teaming up with defence and industrial partners around Europe in an ambitious plan to combine the power of a number of simulations for modelling scenarios where infrastructure fails. The project is called the Design of an Interoperable European Federated Simulation Network for Critical Infrastructures. Yes, quite a mouthful indeed. But one of the key people involved here at Imperial is Erol Gelenbe. He’s a professor in communication systems and networks and holds the Dennis Gabor Chair in the Department of Electrical Engineering.
Erol Gelenbe: We realised that a lot of these so called critical infrastructures are interconnected. For instance, a transportation system also relies on communications, just like a communication system will rely on electricity. And electrical system will rely generally on the availability of other forms of energy such as gas or coal or wind power or sea power.
GM: And I guess one really good example of an infrastructure that relies on other infrastructure is the electricity power system obviously. And the effect, for instance, that flooding has. I wonder if you can talk me through that because there are so many interconnected factors, I suppose, when one of those things fails?
EG: Yes, of course. We realise that a lot of the other critical infrastructures are buried now. So they’re underground in more or less watertight areas but only more or less so because of the amount of water that can damage anything, as we know. So if we have, for instance, flooding, climatic change with a lot of water coming down on us and we’re getting water rising somewhere of course there are pumps that get activated. These pumps of course rely on electricity. Now, electricity is coming through lines underground and when those lines get very, very wet you’d expect that at some point the electricity system will fail and at that point the pumps will not be working as properly as before and, of course, then we’ll get even more flooding. So you see how all these things are highly connected.
And of course if you have flooding you want to evacuate people. You’re relying on a transportation system which is above ground but which may be under water. So all of a sudden your other critical infrastructure may not be working because of the flooding. So again you have these interactions between different important infrastructures. And understanding what would happen, what would be the chain reactions in these systems, is really the subject of our project.
GM: It really is like a big but very serious game of What If? So what if we have a certain of rainfall? What if the water rises to this level? What if then the electricity fails. And so on.
EG: Yes, it’s what if and that leads to simulation. Of course we don’t want to find out about these things when they actually happen. We’d rather be able to predict what might happen so that we can take proper measures or proper precautions to mitigate or to reduce or even to eliminate the effects of these cascaded reactions. So the idea then is to have a set of simulators each perhaps specialised in a domain or a in a sub-domain. For instance, part of a telecommunication system or all of the rail transportation system or part of the other ground transportation systems we can think about, or the electric power supply system and so on. So you’d have a serious of simulators and these simulators would have to interact with each other. I mean, just running them by themselves is not interesting because you don’t get the cascaded effects. So we would be able to run these simulators together in a federated manner as a facility for anyone interested in understanding these chain reactions to use and test.
GM: Where are we at, at the moment? Do we already have some of these simulations and this project is really about bringing them together? Is that pretty much where it’s going at the moment?
EG: Well, the generation or the development of simulators is an unending task. Why unending? Because, for instance, as computers become more powerful, as programming languages become better, we get better simulators. We are able to simulate more complex systems. Also, as we understand the physical phenomena that underlie the various effects we’re talking about. For instance, meteorology is an area where we have increasingly better models of a physical phenomenon and increasingly better predictions. So we have a whole set of simulators which exist which depend on our understanding of the underlying physics, which also depend on the quality of the computers, the capabilities of the computers and the programmin g languages that we use. So the simulators are constantly evolving and changing. Our role is not to develop specific simulators but to allow them to interoperate and work to gether and to be available to qualified users.
GM: Thinking about different scenarios then. For instance, supposing I was a hospital could I come to you once this project is up and running and say I really need to have contingencies as to what happens if the power in the hospital goes down for an hour or maybe two hours or three hours? That’s the kind of project I could bring to you and you could give me some answers about that?
EG: What would happen would be the domain expert from the hospital would come over and access a portal. So the access would be a portal. And the centre that we’re designing or preparing is going to be necessarily virtual and distributed. It’s not going to be localised in one single place. This centre itself will be a critical infrastructure so we don’t want it to rely on just one place.
GM: It reminds me a bit of the grid computing approach, you know, to the Large Hadron Collider at Cern obviously coming on stream this year. It’s huge amounts of data to process and it’s not going to be processed in just one place but in a distributed way through this idea of grid computing. Are there parallels here?
EG: There are definitely parallels but there are also some major differences. For one thing, we’re dealing with an activity which has a large commercial domain interface. The simulators themselves are going to be proprietary. They’re not like public open research pieces of equipment. The simulators are proprietary. The problems have confidentiality requirements. Most of the potential users will not want it to be public knowledge that, for instance, they are even looking at these things because it might impair their credibility. So we have a lot of real world constraints related to this.
GM: Where are we going now? Clearly the project is underway. The funding is there. When are we actually going to see this portal, this resource, for people who need simulations of these kinds?
EG: So the requirements and specifications are being developed as we speak. Some of them are being developed in offices on this floor here at Imperial College. For instance, everything that’s related to the communication network infrastructure for this system that we’re imagining, that we’re envisioning, all of that is being developed by our group. So this kind of work is going on and by the end of 2009 we should have a blueprint for this centre. The business model includes the economics, the legal aspects, the confidentiality questions, the performance issues and so on. So all of those things will be available as a report, as a global report, at the end of 2009.
GM: Professor Erol Gelenbe speaking to me there in the Electrical Engineering Department. In a moment another critical infrastructure, the internet and what’s in store for its future and the way it affects all our lives. But before that let’s have a few of the stories that have put Imperial in the news headlines recently.
Oil and water actually like each other more than we originally thought. So say researchers in the Chemistry Department here writing in Physical Review Letters. It’s always been thought that when water and oil meet they’re separated by an layer of invisible water vapour. The water/oil boundary is fiendishly hard to study as it fluctuates with the movement of the two liquids. So the scientists modelled the boundary on a computer and surprisingly it turns out that even thought they don’t actually mix water and oil do come into full contact with each other in a strong rigid boundary. The insight from this slick piece of research should help scientists and engineers working on oil based products like detergents. And there may even be applications in human biology and medicine as ourselves are formed of watery contents bounded by walls made of oily membranes.
And staying with medicine, and experimental treatment for early stage Alzheimer’s disease, studied by Imperial researchers, generated quite a few headlines around the world recently. The team from Imperial’s Division of Neurosciences and Mental Health, along with colleagues in Australia and Sweden, was looking at a drug called PBT2 which counteracts a protein associated with the disease called amyloid beta. People in the early stages of Alzheimer’s have impaired function when it comes to so called executive tasks like ordering information or planning a sequence of events. But patients given this experimental drug called PBT2 had improved performance in such activities compared to those given the placebo. The drug proved safe and well tolerated, although writing in the journal Lancet Neurology the researchers do point out that this was a relatively small trial. The results are encouraging but larger follow up studies are now needed.
And of course you can catch up on more news and events from around the College at imperial.ac.uk/news.
Well, there’s a bit of hubbub around me now with the clanking of plates and the clinking of cups. That’s because I’m at a meeting. This particular one is called the Wealth of Networks and one of the organisers is John Darlington. He’s the Director of the Internet Centre here at Imperial. What’s the idea then John? The Wealth of Networks, why the conference and why now?
John Darlington: Computing and the internet has had dramatic effects on people’s lives and productivity of industry and the way government is run but there’s a feeling that we’re due for another revolution with the development of broadband, the convergence of mobiles, with the internet that there could be a next step up where the internet could become much more integrated with our lives. Much more personalised. Much more helpful and transformational in the way we do things.
GM: I suppose it’s all about seeing where we go from here. Because already so much of our lives in terms of commerce are on line. People tend not to go to the travel agents on the High Street to buy their holiday. They book that on line and flights and so on. So we’re already in this incredible place. One that we might not have foreseen maybe as recently as 10 years ago. You’re seeing where next?
JD: Well, one of the problems of the internet is you can never quite see where next. Suddenly something happens and you find yourself thinking that was obvious. Why didn’t we think of that before? But the internet has become very useful; downloading music, socialising and so on. But there are perhaps some slight clouds on the horizon. The record industry is still fighting against free downloading. There have been some negative aspects of social networking, bullying and so on. So one of the things is how can we take the internet forward and maximise its benefits but perhaps erase some of these anomalies?
And some of the work in the economics we’re doing is looking at business models that actually encourage the downloading and sharing of music but still the end generates some financial reward back to the original creators. Ditto in the social networking. The sociologists and the mathematicians are looking at the patterns of network interaction and how they evolve. Why does Wikipedia generally produce very good material and how does the bad material get spotted? And can we apply that in a more general principle so there can be social organisation on the internet not regula tion which encourages good behaviour and gets the maximum benefit for the society, business and commerce.
GM: I was going to ask you about the regulatory side of things. So thinking about the music industry, how you make sure that artists don’t get ripped off. Wikipedia, making sure that the content is good and people don’t vandalise it with propaganda or malicious articles. Is regulation part of the answer there or do you think market forces are the answer? How do you kind of marry together market forces and regulations?
JD: I think a combination of the both. I think the one message is that the internet has introduced new patterns of distribution of material. Those goods have inherently copyable properties. In economics they’re called public goods. Ditto the internet has allowed the social interaction. So there’s no way of stopping that. Any attempt at regulation that says people can’t download and share music or people can’t communicate in networks just goes so much against the grain that it just won’t happen and people will just ignore it. Or the regulation will be so heavy it will kill off all the good with the bad. So what we’re trying to look at are business models that go with the grain.
That enables people to do what they want to do naturally but, as you say, also ensure that end producers of digital content do get some reward proportionate to what they do. And the models we’re developing actually mean that more revenue can be got. Because if you encourage copying more people have access and use the material and experience it so in the end you would think there’d be more revenue streams back to that. So the idea is to produce economic and social models that enhance this and then see if there’s some perhaps some minimum of regulation that would enable these things to work properly to try and avoid the dead hand of saying it’s all going wrong and we’ve got to clamp down on everything, which would throw the baby out with the bath water.
GM: And you’ve called this meeting the Wealth of Networks. Obviously borrowing from the Wealth of Nations, the well known economic text by Adam Smith from the late 18th century. What do you think he would make of all this?
JD: I think he would love it actually, to be honest. One of the observations is that the internet approximates the idealised free market much more closely than the real world does. It’s instant communication, there’s very low barriers of entry and there’s much closer to perfect knowledge. So one of the observations is that the hidden hand, the open market, could work much better on the internet than it does in the real world. Because (a) on the internet consumers have much greater power, you know, look at uswitch and confused.com. Much more information, much more power. And monopolies can be circumvented because of the low barriers to entry. So I think he would be really delighted and see how the hidden hand is working on the internet so bring maximum social and economic benefits. I think he would perhaps agree that some light touch of regulation is needed to make the market work efficiently but not to stop people doing what they naturally do.
GM: Professor John Darlington at the Wealth of Networks meeting recently held on our main campus here at Imperial.
Finally though, to the West London Renal Health Centre which is part of the Imperial College Healthcare NHS Trust. It’s the busiest transplant centre in the country carrying out 200 procedures every year. Organ transplantation is a fast moving exciting area of medicine. New ideas in the subject always receive attention in the medical journals and the wider media. But alongside these new scientific insights in a radio assignment as part of her Science Communication MSc here at Imperial Georgia Gale Grant was also interested in hearing the voices and personal insights of those involved. So she interviews two physicians in the Renal Health Centre and one of the patients. And keen to innovate in this exercise here Georgia has adopted an experimental approach to her feature making combining her interviews with poetry. The editing style is unconventional. The rapid and frequent transplanting of the voices intended to reflect the subject matter. So here is Georgia’s short feature for radio.
Georgia Gale Grant: The Gift by Karen Weddeck. I received a gift back in 72. It came from someone I never knew.
Nadey Hakim: I always say that the kidney is priceless.
GGG: It came with no card, ribbon nor bows but sent to me, that’s all I know.
NH: You can’t create it yourself. The dialysis machine is an artificial kidney.
GGG: It didn’t arrive in a box or a sack and is a gift one would never send back.
NH: But it cannot beat what a kidney is.
GGG: The gift in size is rather small compared to its power to conquer all.
Anthony Warrens: I’m Anthony Warrens. On the academic side I’m a reader in immunology and renal medicine at Imperial and I’m a consultant physician at the renal unit at the Hammersmith Hospital.
Nadey Hakim: I’m Nadey Hakim, Surgical Director of the Transplant Unit at the Imperial College Healthcare NHS Trust. I’m in charge of the surgical aspects of the unit.
AW: The best form of treatment for somebody who’s kidneys have failed is transplantation. It approximately doubles their life expectancy and improves their quality of life enormously.
Patient: Before the transplant I was on dialysis. I was being monitored and we were hoping to go from being monitored to a transplant. Unfortunately I had to do a bit of dialysis in between.
AW: There is good evidence that actually the best thing to do with somebody whose kidneys are failing is to try and predict when the point of failure occurs and to transplant them at that point.
NH: That’s what we call pre-emptive transplantation. It’s a much better option than to have somebody who starts dialysis and gets depressed about dialysis.
Patient: The hardest part was trying to ensure that it didn’t overtake my life.
NH: And this is less likely to be a nice environment for the patient.
AW: There are problems with that. Number one, it’s actually difficult to say exactly when the kidneys are going to stop working. Number two, there’s not an inexhaustible and easily accessible supply of organs. There are two sources of organs. One is from people who have died. There’s a great shortage of those. The second is a friend or relative of the person with renal failure.
NH: A live donor transplant is a much more successful transplant. Because if you think about it the organ first of all is going to be much better matched. Second, it’s going to be much more fresh.
AW: By definition the person giving the kidney is in better health and therefore the prognosis for that kidney is better.
NH: Having a cadaveric donor where the kidney is put in ice for several hours if not for a day or two. So in other words the live donation is so much quicker. The kidney remains out of the human body for only half an hour or an hour and is alive again. So it’s bound to work quicker, better and last longer.
Patient: My brother offered straightaway. By the time we’d done the tests and everything he was a very suitable match for me.
NH: The improvement in the patient is absolutely immediate. Within 24 hours they’ve changed their features. They’re usually pale looking, anaemic etc., and then as soon as you do the transplant they’re back to as healthy as you and me.
AW: We see patients after a transplant very frequently.
NH: This is where I think we have done better than most centres is the fact that we have such an active follow up programme with so many doctors following up the patients.
Patient: It’s a learning curve at the moment because I need t o find the right balance of tablets and ensure that I’m okay with taking them and the side effects are limited.
AW: One of the nice things and what’s attractive to me is you can really make a major difference to people’s quality of life and indeed their life expectancy. And that’s obviously very satisfying as a doctor.
GGG: You’ll never hold this gift in your hand but it’s given to people throughout the land.
Patient: The first thing I noticed was not being so tired. I was only tired because I didn’t sleep very well in hospital.
GGG: This beautiful gift has set me free. What a precious gift and given to me.
Patient: They are such a fantastic team. You just can’t get any better.
NH: I believe that regenerative medicine is undoubtedly going to be the future.
AW: I think stem cells are fascinating. I think the potential for repair and regeneration of organs is enormous. And if we learn the biology of stem cells in may very well be that I’ll do myself out of a job and we won’t need to transplant other people’s organs. We’ll just grow back our own.
NH: Whether it is going to happen quickly or not is the big question. But certainly in the next one or two decades.
GH: Professor Nadey Hakim of the Transplant Unit at St Mary’s ending that feature from Georgia Gale Grant. And you heard the voice of the patient who gave us permission to include her in that piece although she prefers not to be named. Also speaking was Dr Anthony Warrens of our Hammersmith campus. And that pretty much wraps it up for this month. A quick apology though before I go. Last time I promised you a ride in our very own flight simulator. Unfortunately that piece has been delayed due to adverse weather conditions and the late arrival of the incoming aircraft for today’s flight. I’ll do my best to bring it to you next month instead. But I can promise that I’ll have a packed pod for you in October. As with all editions of this podcast it’ll be available on the first working day of the month. I’m also keeping the promise I made to Ozgur Buldum to name check him as the composer of our theme music. And to our press office who, along with the Science Communication group here at Imperial, collaborate to make this podcast possible. So have a great September and I’ll see you next month. Thanks very much for listening and goodbye for now.