The historic COP26 Summit in Glasgow sparked big, bold statements by political and business leaders about their commitment to taking action, but the real challenge will be the doing.
How will companies keep their promises and actually achieve them? With all the different measures that can reduce carbon emissions, it may not seem obvious but data skills can play a key role in accelerating ambitious sustainability targets and helping to save the planet.
Many businesses nowadays have incredibly ambitious sustainability targets. Take Burberry, for example, which has pledged to cut emissions across its extended supply chain by 46 percent by 2030 and committed to being carbon positive by 2040. Meanwhile, Unilever has set out to eliminate direct greenhouse gas emissions from its operations and is aiming to achieve net-zero emissions from its products up to the point of sale by 2039. But big businesses such as these won’t sacrifice profit, so it’s vital that sustainable business practices offer both commercial and environmental benefits.
From my role working at Decoded, which trains business leaders and corporate executives on their data and digital skills, I have seen countless examples of how data, artificial intelligence and machine learning can make a company not only more efficient but also more sustainable.
Take Google DeepMind’s almost folkloric story. Back in 2016, DeepMind set out to reduce the amount of energy required to cool Google’s data centers. For context, the technology company uses around 12.4 terawatt-hours of electricity per year, the equivalent of powering over 3.3 million homes in the U.K. for a year.
DeepMind’s all-purpose algorithm subsequently devised a real-time, adaptive system that cut the cost of cooling by 40 percent and the overall energy consumption by 15 percent. This both saved Google a significant amount of money and reduced its environmental impact. (Google has set out to operate on carbon-free energy 24/7 by 2030.)
Data skills can transform manual processes into automated ones, leading to huge efficiencies, ensuring that employees can focus their time on truly impactful work.
It isn’t just the likes of tech titans such as Google that are ramping up their employees’ data skills. For shipping and logistics businesses such as UPS, more efficient delivery routes save drivers time, reduce fuel use and ultimately increase customer satisfaction. This is especially the case at a time when more packages are being delivered than ever before due to an increase in online shopping as a result of the pandemic.
In the second quarter of 2020, for example, UPS delivered over 21 million packages on average every day, a 22.8 percent increase on the year before. Shaving off just one mile for each of its drivers per day could save the company up to $50 million a year, and this is where machine learning steps in. By using a proprietary tool called Orion, which uses advanced algorithms to create optimal routes for delivery drivers from the data supplied by customers, drivers can alter their routes on the fly based on changing weather conditions or accidents. UPS estimates this insight could reduce delivery miles by 100 million per year, the equivalent reduction in carbon emissions as taking 21,000 passenger cars off the road for a year.
Data skills can transform manual processes into automated ones, leading to huge efficiencies, ensuring that employees can focus their time on truly impactful work.
A relevant case study from our work at Decoded was with a retailer, whereby an employee set out to automate a process previously carried out in Excel that took 20 hours weekly to complete. By creating an automated process using database and programming tools, the business was able to reduce a process from 20 minutes to just 10 minutes as well as create something that could be used business-wide.
Using data skills enabled the retailer to optimize the process, meaning more goods could be shipped through the distribution center. This led to more efficient shipping which has the ability to create a CO2 saving of over 1.2 million metric tons per season — the equivalent of more than 2 million people flying from London to New York.
The reality is that that employee didn’t set out with the goal of reducing CO2 emissions with their new data skills. They simply wanted to automate a very manual process and save their team time; the environmental benefit was a very positive side effect of their original intention.
Using data skills and machine learning can uncover inefficiencies in processes that humans would never spot or would take a significant amount of time to find. Data analytics and AI can mitigate human error, meaning tasks don’t need to be redone, saving energy in the process. They can also spot opportunities to use less raw materials. It’s important to acknowledge it can take a significant amount of energy to train machine learning algorithms, but over time the energy savings should outweigh this.
It’s important to acknowledge how far the business world has come in moving environmental responsibility to the mainstream. But the urgency of this moment means corporate leaders must back their commitments with genuine action and ensure their people are given the tools and skills to achieve these ambitious goals. As corporations move to accelerate action and implementation after the pivotal COP26 talks, it’s time to realize the commercial and environmental potential of machine learning and data skills.
