Sustainability – the ‘smart’ way
Humans have been obsessed with the idea of artificial intelligence for centuries. C3PO, J.A.R.V.I.S., the Matrix, Tin Man in The Wizard of Oz – these fictional computers and robots, among others, have featured heavily in some of the biggest films of the 20th and 21st centuries. Over time, many have envisaged how they believed AI would work in the world of the future, some picturing clunky robot chauffeurs, others focusing on forbidden loves and friendships. However, one element of AI that hasn’t been considered by Hollywood, perhaps due to it not being a particularly dramatic box office draw, is its potential role in combatting climate change. How? Simple: promoting efficiency.
The world of AI has an up and coming superstar in the public consciousness. Self-driving cars are set to hit the public market over the next decade or so with recent trials from Apple, Waymo (formerly the Google self-driving car project), and Tesla. But these autonomous vehicles haven’t just played a huge role in the plot of Incredibles 2: if steered in the right direction, they could play a role in reducing our emissions. With more cars than ever on the roads, private transport continues to be a leading source of pollution; combining self-driving cars with the transition to electric vehicles could significantly reduce carbon emissions: fuel consumption could also be decreased with more efficient route algorithms, communicating with other vehicles on the road to ease congestion, something all non-fresher Warwick students would be grateful for.
While there exist very valid concerns about data privacy, cyberterrorism, and machine ethics, it is also possible that, without careful policy strategy, autonomous cars could encourage more driving, outweighing their potential co-benefits in the form of carpooling and route planning. It is critical that over the next few years, attention is paid to how these new technologies will be rolled out to maximise the environmental benefit of self-driving cars.
In the flurry of excitement surrounding electric vehicle transition, hybrids and autonomous vehicles, another consequence that is frequently overlooked is the strain on our electrical grid. The power cut in August which left more than 1 million people without power highlights the lack of resilience in the UK’s National Grid as we continue to use more and more energy. This is where ‘smart grids’ come in to play.
What is a smart grid? The word smart is used in front of a lot of dubious things nowadays (smart toaster, anyone?), but the ‘smartness’ of this idea can be shown through a quick analogy. Take breathing: all humans need to breathe oxygen. There are times when you need to breathe more than others, like during a run, as you need to get the oxygen into your system faster to meet the demand of your body. However, if you breathe too much when your body doesn’t need the extra oxygen, like when you’re sitting down, you run the risk of hyperventilating.
Our brain and the rest of the body communicate in order to match the supply and demand of oxygen and they generally do a pretty good job.
Imagine if instead there was a machine controlling our breathing. And this machine would only allow a fixed, constant amount of oxygen to reach our lungs. This would be fine for normal everyday tasks but as soon as we reached the extremes of our demand, such as running or sleeping, it could cause some significant problems. It would not be a beneficial system.
The latter scenario is akin to a simplified version of much of our modern electricity grid. And the former represents smart grids, and in particular what can be achieved with the introduction of AI.
Smart grids allow for two-way communication between the supply and the demand, producing a better understanding of the two and therefore a more efficient system, as well as paving the way for more flexible tariffs based on current demand. We take for granted the huge amounts of data our own brains process in order to breathe, but to manage the amount of data required to produce a smart grid, we need artificial intelligence. AI can also learn from its choices and mistakes, adding further to the value of smart grids. The inclusion of this technology within our energy infrastructure would provide greater resilience, alleviation of grid constraints, better efficiency and save the customer money.
A prime example of smart grid potential lies with our earlier friends, electric self-driving cars, and their expected global uptake. It makes sense to take advantage of the fact that most cars are stationary approximately 90% of the day: in the case of electric vehicles, we can thus treat them as battery storage. This technique is called V2G (Vehicle-To-Grid) and works by charging cars in periods of low demand when electricity tariffs are cheap, and then discharging that stored electricity back into the grid in periods of high demand – providing the vehicle is plugged in. Algorithms are used to determine the optimal time for battery charge and discharge, cutting energy costs to the customer. Currently only a few cars are compatible, but this number is expected to increase as trial projects, such as that planned in Solihull, become more mainstream.
The beginnings of smart grids, at least in a rudimentary form, have already begun to appear in many UK homes: consider smart meters. These unobtrusive devices display your energy usage in real-time, and can help shape consumer behaviour; seeing the real time cost of turning on the kettle can make us think twice. While not part of the past’s artificial intelligence dream, these meters are a stepping-stone towards the adoption of technology to help us use energy more efficiently.
There is an extraordinary range of technologies and methods that fall under the umbrella of artificial intelligence, from the more unusual and abstract to the small and the mundane. Each can be in our toolbox as we try to combat climate change, but only if we use the right tool at the right way – using a saw to tighten a screw is not going to help. The examples above also demonstrate that tackling climate change does not have to come at the cost of productivity or technological advancement. Ultimately, technology of the future should be compatible with its preservation.