Like most people right now, I’m sitting at home as I write this. For many, regular life may feel like it’s on pause but outside nature is carrying on. In fact, while we’re all staying indoors, animals have been quietly reclaiming cities and towns. You’ve probably seen the photos of mountain goats running amok on the streets of a Welsh town, or the critically endangered big cat spotted prowling deserted streets in India. For me, those images are a stark reminder of our encroachment on the natural world.
Though for the time-being it may have taken a backseat in our minds, the climate emergency hasn’t lost its urgency. It may feel like a lifetime ago now, but it was only a few months ago that terrible floods swept over the city of Venice, causing around €1 billion in damage to homes, businesses and historic sites. And that Australian bushfires scorched 5.5 million hectares of land in New South Wales, devastating local wildlife populations.
This is a topic particularly close to my heart. Before becoming a data scientist I worked as an Oceanographer, so I’ve studied the detrimental effects of climate change on natural ecosystems. Just look up images of the Great Barrier Reef from a few decades ago compared to today to see how acidic waters (caused by increased carbon dioxide in the atmosphere) have eroded the hitherto vibrant and thriving coral reef.
Back in 2018, scientists told us we have 12 years to keep the planet’s temperature at a maximum of 1.5C and that just half a degree rise would increase the risk of floods, droughts and extreme heat. Now, there’s a growing consensus that we have something closer to 18 months. Meanwhile the UN has warned that on our current trajectory, we’re heading towards 3C of global warming – which would trigger an irreversible rise in sea-levels and disaster for major coastal cities all over the world.
Yet right now, with less people travelling and economies slowing, some cities and regions are actually reporting significant drops in carbon emissions and many experts expect that the current shutdowns will have an impact on overall CO2 levels for the whole year. But of course, this won’t last forever and the current dramatic measures aren’t the way forward. Lasting solutions still need to be found.
In December 2019, the UN hosted their annual climate talks, which dragged on to be the longest in history. There seemed to be one clear takeaway to me: there’s still a huge amount of uncertainty about the best course of action to slow the warming of our planet.
There is no silver bullet. But many in the tech industry – myself included – believe artificial intelligence techniques have huge potential to help companies reduce their impact on the environment.
A manmade solution to a manmade problem
One of the main ways we’re using AI today is to drive operational efficiencies. Organisations are data-generating beasts and they’re starting to feed that data into algorithms to find where they can streamline processes and drive down costs. But the same technology can also be used to discover new – sometimes surprisingly simple – ways to reduce carbon emissions. Oftentimes, just a small adjustment to the way a company is doing one particular task can have a big knock-on effect on their environmental impact.
There aren’t really any off-the-shelf solutions right now, so businesses must start out by looking at their most emission intensive operations – such as their supply chain, for example – and work from there to consider how AI techniques can be applied to optimise processes.
Let me give you an example. Linage Logistics is a cold storage company that handles frozen goods before they reach supermarket shelves. Keeping food at a consistent temperature of around minus 20 degrees meant their yearly energy consumption was about the same as a mid-sized U.S. city.
But by overcooling their warehouses at night they were able to keep their freezers turned off during the day (for anywhere between 5-10 hours) until the temperature warmed back up, reducing their energy consumption by 34 per cent over the course of three years. Pulling that off took absolute precision only made possible by AI, which could carefully maintain the freezing schedules and predict how external influences like the weather would impact temperatures.
The logistics and transportation industries are responsible for around a quarter of Europe’s entire greenhouse gas emissions – so it’s unsurprising that we’re currently seeing a drop in CO2 levels with less people travelling back and forth. Many of these companies have already been under pressure from increasingly ambitious sustainability goals, but AI could help them achieve and even exceed these.
One such company is Stena Line, one of the largest ferry operators in the world, who approached us to help them reduce their fuel consumption by 2.5 per cent per nautical mile annually. Working together, we developed an onboard AI captain that was able to simulate different scenarios and predict the optimal route for saving fuel. The human captain was able to input various factors – such as wind conditions, water depth and sea currents – and, in real-time, the AI would be able to offer helpful recommendations to keep the ship on the best course.
In both the above examples, the companies involved didn’t reinvent the wheel or completely overhaul their business model. A more sustainable solution was right there in plain sight, but it could only be reached with the help of technology.
Committing to a sustainable future
Research by PwC suggests that deploying AI solutions could reduce worldwide greenhouse gas emissions by 4 per cent in 2030 – the equivalent to the annual emissions of Australia, Canada and Japan combined. But pulling that off will take a combined commitment from both businesses and governments to invest in the right technology and infrastructure.
Here’s a flavour for what that could look like. Since 2015, the Isles of Scilly – one of the most protected landscapes in England – has been the site of an ambitious project aimed at reducing the islands’ carbon output (ranked the 8th highest for fuel poverty in all of England) with technology.
Hitachi, the Council of the Isles of Scilly, the Duchy of Cornwall, Tresco and the Islands' Partnership joined forces to reduce the carbon footprint of the island and optimise locally-produced, renewable energy. The project leverages around 400kW of solar panels installed on 70 homes across the island, partnered with Hitachi’s IoT infrastructure and AI that is capable of learning energy consumption habits and optimising how power is collected and used throughout the home.
Hitachi is very much on the frontline helping cities like London to reach their sustainability goals. In 2019, we co-launched (alongside Ofgem, UK Power Networks, Royal Mail, Centrica, Uber and Scottish and Southern Electricity Networks) the world’s largest commercial EV trial. Optimise Prime will help the project partners devise practical ways of overcoming the up-front costs that are currently preventing widespread commercial EV deployment, while reducing the cost of the EV transition for electricity bill payers. The dataset will be publicly available, allowing urban planners, power grid engineers and, of course, vehicle operators, to prepare for EVs.
That may sound somewhat fanciful now, but we must challenge ourselves with ambitious goals. The technology is here, and we’re seeing it explored by some businesses already, but there needs to be more wide-scale commitments to leverage technology for the preservation of the natural world.
Dr Anya Rumyantseva, Data Scientist, Hitachi Vantara