Artificial intelligence will soon be helping smart meters run your home, reports E&T.
In the race to develop a UK smart grid, the Department of Energy and Climate Change has stated that by 2020 every single UK home will have a smart meter. As industry debates the best way to roll out the next generation of electricity meters on such a massive scale, computer scientists are busy developing software to help you make more of your meter.
Professor Nick Jennings from the University of Southampton, UK, is one such scientist. While a basic smart meter provides real-time communication between a household and the grid so you can tailor your energy use according to supply, Jennings and his team are writing software programs that do this for you. As he puts it: 'We come from an artificial-intelligence background and we're interested in getting smart meters to do clever and useful things.'
The researchers are currently developing 'home management agents' for smart meters. These programs will reside on the meter and optimise the way you use energy in and around your home, hopefully saving you money at the same time.
'An agent will work for you, figuring out what you should do with your energy. Should you use it, sell it back to the grid or store it?' explains Jennings. 'You may want to defer your load by charging up, say, a battery or your plug-in hybrid vehicle when electricity is cheaper and then use it when electricity is more expensive, so you'll end up saving money. An agent can [organise] this for you.'
According to Jennings, the agents will also learn about you and the way you use energy in your home, without being told. And, by combining this knowledge with local pieces of information picked up from external sensors or the Internet, it will independently optimise energy use in your home (see 'Neighbourhood watch' box).
'I think within a year we're going to have a concept demonstration of an agent that can learn the pattern of your life, learn home management capability and then make decisions,' says Jennings. 'For example, you might have a temperature sensor outside your home or maybe the agent could look on the web for a local weather forecast. It would then tailor this information to you, your house and manage the load around your house.'
But how long will it be until we actually see this 'smarter' meter in our home? According to Jennings, soon.
'From a technology point of view, we can already implement these on smart meters, and plan to do so in the next couple of years. We've collaborated with a smart-meter manufacturer that intends to put the technologies we develop on its suite of smart-meter programs,' he asserts. 'However, this does also depend on the roll-out of smart meters and what happens in the marketplace is a bit beyond my control.'
As part of the research, the team has also used computer modelling to understand how energy storage in the home will affect the consumer and the grid. For example, recent work uses elements of game theory to investigate exactly how the home management agent should manage and store energy to maximise profit for the consumer and benefit society as a whole.
The team worked on the principle that the agent would minimise costs for a consumer by storing energy when prices were low and using that energy when prices were high. It would also learn to alter how much energy is stored according to market trends such as fluctuating energy prices.
Whilst the researchers found that both the individual and society profited from the use of home energy storage, they also discovered that only 38 per cent of the UK population would need such a storage device for everyone to reap the benefits. Indeed, at this penetration, the researchers calculated that consumers would, on average, save 8.5 per cent from their annual electricity bill, which is £60 based on a typical electricity bill of £675. At this point the overall decrease in carbon emissions was at its greatest - 7 per cent.
'We set out thinking, okay, the more storage everyone has, the better off everyone would be. This isn't true,' says Jennings. 'You might want to quibble over the exact number - there's always different bits of modelling you can do - but the interesting thing is the optimum level isn't 100 per cent, it's a long way from 100 per cent.'
Jennings believes this spells good news for the power industry. 'Not everyone will be able to afford or be interested in storage capability, but everyone will benefit assuming a reasonable number buy-in,' he says.
The researchers now intend to work out what happens when a consumer asks his or her smart meter to favour renewable generation, rather than the cheapest energy. As Jennings points out, some energy consumers are simply 'price-point conscious', whereas others are also very concerned about the environment and will want to buy the 'greenest' electricity.
With this in mind, the Southampton team, led by Jennings's colleague Dr Alex Rogers, has already developed an iPhone application called GridCarbon, that measures the carbon intensity of the UK grid on a half hourly basis. Launched earlier this year, the application has already been downloaded more than 1,000 times and the researchers have just completed a version for Google Android, the Internet search engine's operating system for mobile devices.
'This is the sort of information that can be fed into your agent,' says Jennings. 'Home management agents are a huge area of research, and it's an area in which we're getting very much involved.'