On 25th June 2103 I attended this presentation by Andy Hopper, President of the Institute of Engineering and Technology (IET) and director of the Cambridge University Computer Laboratory, as part of the York Festival of Ideas. This is what I learnt.
It’s about the crossover between computing and sustainability. It’s not the answer to sustainability but it is a contribution to the answer to sustainability.
Green computing is about establishing an optimal digital infrastructure. For example, 3-4% of the world’s total energy is used by computers and communications. How much of this energy consumption can be offset? Prof Hopper described how technology can be used to move the processing load around the world to where surplus energy is available. Renewable energy is notorious for intermittence. Sometimes power is generated but there is no use for it so it is wasted. If data centres are located close to sources of renewable energy these surpluses can be used. Global networks permit processing to be carried out wherever this “free” energy is available. Google, which alone has some 1.5 million servers, is already using software to carry this out automatically.
Data centres can be designed for economy and efficiency and computer architecture can be built so that instead of having two extreme states – standby and full power – there can be a proportionate increase or decrease in computer usage as processing demand fluctuates. Increasing hardware performance and more interconnected devices will improve energy efficiency. Workload trends will also affect energy consumption - increases in batch processing and increasing amounts of data produced and consumed. Recognising that not all processing is needed on demand but that some can be scheduled to smooth peaks and troughs will also improve efficiency.
All these measures could lead to future energy savings of 15-20%.
Computing for Green
Having established the most green and efficient computing infrastructure, how do we use this to green the rest of society? Professor Hopper talked of a “Google of Everything” - a universal data resource available to all and with information about everyone and everything. For example, Google Street View could be adapted to give an infra-red view and to immediately reveal where houses are losing heat. This is a first step to saving energy on heating. He told us about Ubisense, a system for tracking objects in three dimensions with an accuracy of 15cm. This is used by BMW not only to track assembly equipment, but also to programme it according to its next task. As the unit approaches a nut it is programmed with the number of turns it should give and the required torque. As it moves to the next nut the system automatically re-programmes it. The result is increased efficiency - more accuracy, less manual intervention, time saved. Accurate tracking already saves money and improves efficiency for Airbus, in public transport depots, dairy farms, distribution centres and convalescent homes.
The big question is how far we go to track ourselves. It is possible to track an individual’s energy footprint, carbon footprint or water footprint. It is also theoretically possible to track an individual’s location to a very high degree of accuracy. With data as detailed as this we can take action to control the use of energy, carbon or water. This would include managing services within buildings according to their occupation and use, monitored in real time. We can give people feedback and tailored incentives. Other sorts of wearable sensors could monitor an individual’s state of health, so if they felt unwell they could go to the doctor and provide a read-out to help him make a diagnosis. (Or maybe the sensor detects a life-threatening condition and automatically calls the ambulance!) A major obstacle to this is the trust issue. Will people believe that their privacy is respected and trust their governments with their data? Just at the moment, with the GCHQ and PRISM issues, evidence of cover-ups and secrecy in the NHS and the police and a legal system way behind technological reality, many people will refuse to participate.
Another example of the benefits of technology is the electronic shoe. A shoe with strain gauges built in to analyse the gait of the runner accelerating, cruising and slowing down. Valuable information for trainers, but an adaptation could monitor the runner’s weight and send an alarm if it fell out side the normal range. Raspberry Pi, the ultra-cheap computer, gives today’s youth an introduction to computing and makes them aware of the threats to privacy from surveillance and hackers. At the same time, this generation is totally relaxed about posting almost anything on Facebook - the bad and the good for all to see for all time. Smartphones, particularly with location services enabled, potentially compromise privacy. However, it is now possible to use fake data when requested by apps, to get the service without giving away private details. But how long before this “mocking” can be detected?
Will we demand privacy and restrict data which could otherwise improve sustainability - consumption patterns, energy use etc? Trends indicate that we will allow our privacy to be gradually eroded. After all, caller ID was once thought to be the work of the devil and now everyone takes it for granted.
Computer systems are now crucial to society. They are like a pacemaker for the planet: we cannot afford them to stop. We can use more detailed knowledge of ourselves and the way we use the world to take measures to tackle global warming, traffic management and energy demand. However, the more data we manage, the more challenges arise. What parameters of error can we accept? Is there a clear audit trail to authenticate the data? How can we assure privacy and security?
For the future, Professor Hopper looked forward to robust, self-correcting energy management and mapping systems on a global scale. He’d like to see a method of absolutely removing an image from the internet. At present, nothing can be revoked. Those party pictures you posted on Facebook are there for ever. He’d like to see an application which analyses the correctness of news. He called it a virtual BBC “More or Less”. [More or Less is a BBC Radio 4 series which picks out statistics in the news and shows how they have been twisted, misquoted or wilfully misunderstood.]
Wealth in Cyberspace
This last section of the presentation was quite difficult to understand. Our speaker mentioned an “Ebay for all”. Does he mean that the internet will provide everything we ever need. That we’ll shop on line, meet up on line and have all our life experiences on line? Will we no longer need to travel? He suggested that the developing world could go straight to this state of data abundance, without going through the stages that the developed world passed through.
Not altogether sure what he meant here. Has he read “The Machine Stops” by E M Forster, I wonder? (Perhaps Ross Noble has read it, although it doesn’t mention moles.)