1 - Climate change presents a global challenge, and AI offers powerful potential to drive rapid and extensive transitions toward sustainability across sectors, including transportation, construction, and retail.
2 - In transportation, AI helps optimize ship timings, route optimizations, and reduce empty truck travel, while in construction, AI supports energy efficiency in buildings, waste reduction, and sustainable material use. Retail AI applications can promote eco-friendly consumption and manage waste.
3 - While AI holds transformative potential, ensuring the quality of training data, understanding model uncertainties, addressing ethical considerations, and considering AI's own environmental footprint are critical.
These three industries must undoubtedly play a critical role in a global transition toward a climate-friendly future. According to the EPA1, transportation accounts for 28% of global carbon emissions. Meanwhile, according to the World Green Building Council2 , buildings account for 39% of global energy-related carbon emissions, with 28% coming from operational emissions and the remaining 11% attributed to the construction process, including the production and transportation of materials. Retail, while often overlooked, is also highly important, because it is ultimately the interface between producers and consumers, and therefore has the power to drive both supply and demand towards sustainable practices.
The development of powerful AI algorithms has come at an opportune time. AI is not just a technological marvel but an opportunity to greatly accelerate the global transition towards sustainability, which is becoming more and more critical. There are several reasons why AI in particular, when used smartly, is well-suited to this task:
1. Pattern Recognition: AI excels at unearthing patterns that can be unintuitive and therefore overlooked by humans, which can lead to impactful emission-reducing strategies that might otherwise be overlooked.
2. Optimization: AI can refine processes, including manufacturing, transportation routes and industrial operations, making them greener and more efficient, greatly outperforming traditional methods in terms of efficiency.
3. Speed, Scale, and Efficiency: As the need to accelerate industry-wide changes to meet climate targets is becoming ever more urgent, AI's ability to generate rapid predictions and optimized processes enables timely interventions that are critical towards meeting these goals.
4. Beyond Human Boundaries: AI excels at processing data and solving tasks that are too intricate or vast for human analysis, thereby expanding the areas in which opportunities for greenhouse gas emissions reductions can be found.
Climate change due to greenhouse gas emissions presents significant challenges to ports worldwide, especially those situated in susceptible coastal regions. According to the University of Oxford’s Environmental Change Institute, a staggering 9 out of every 10 major ports are at risk from windstorms, flooding, and the rising sea levels3. The implications for international trade are alarming, with potential disruptions escalating shipping sector expenses by a projected USD $25 billion each year by the end of the century. Stakeholders in the maritime industry have begun to use AI to reduce GHG emissions in several ways. Some examples include:
- The Port of Rotterdam, renowned as one of the globe's most expansive ports, has launched the "Pronto" initiative, a project that uses AI to forecast the best times for ships to arrive and depart4. The result is a reduction in vessel waiting times by as much as 20%, leading to decreased fuel usage and emissions.
- Wallenius Wilhelmsen has embraced AI to refine their voyage strategies. In a collaborative effort with DeepSea, they've developed AI-driven tools that provide ship captains with intricate route optimization guidance, resulting in up to a 10% cut in fuel consumption5.
The trucking industry, too, is undergoing an AI-driven transformation towards more sustainable practices, with initiatives including:
- Pioneers like FarEye have introduced dynamic routing solutions that promise to curtail miles driven by 8% to 12%6. This is not only a cost-saving measure, but leads to fewer carbon emissions and a cleaner environment. Similarly, Google’s AI route planner has the potential to reduce over 1 million tons of carbon emissions annually7.
- Another innovative use of AI is the optimization of cargo loads. Companies like Flock Freight are leading the charge in this domain8. By leveraging AI to combine shipments from diverse cargo providers, their system enables goods to be transported efficiently within a single truckload. This not only ensures optimal utilization of available space but also significantly reduces greenhouse gas emissions, with reductions ranging from 15-40%.
- A significant concern in the trucking industry has been the 35% of miles that trucks travel empty, leading to unnecessary fuel consumption and increased emissions. Digital freight platforms like Convoy, Uber Freight, and Doft are addressing this issue, by building AI-powered tools that enable trucks to reduce empty travel time, by securing backhauls upon booking a load and combining multiple assignments, ensuring fewer empty return journeys9. The tangible benefit is a marked reduction in unladen travel distances and a corresponding decrease in CO2 emissions.
While technology offers promising solutions, it isn't the sole answer. The need for international cooperation from businesses and governments is more crucial than ever to develop better-coordinated strategies to more effectively combat climate-related challenges. One area where such collaboration can have profound impact is data, which is often isolated among different parties, and difficult to combine into a cohesive view. This makes it challenging to build solutions that can take a more holistic approach to produce greater overall improvements in efficiency in the supply chain. Improving data sharing and standardization across the transportation space is a challenging but important effort to build foundations that can ultimately enable greater reductions in carbon emissions across the industry.
The construction sector is a significant contributor to global carbon emissions. With a majority (80%) of the buildings that will exist in 2050 already constructed, much of the focus shifts to retrofitting and optimizing current structures. According to a study by the IEA, AI has the potential to reduce energy consumption in buildings by up to 40% and subsequently lower carbon emissions10. Companies are harnessing AI to address these challenges:
- The Arup Neuron system is a Digital Twin platform that uses AI to forecast demand for cooling systems that has enabled reductions in building energy efficiency by up to 15%11.
- BrainBox AI offers an AI-driven HVAC optimization solution that uses GHG assessment data to reduce GHG emissions and energy consumption in buildings, resulting in up to 40% lower HVAC-related carbon emissions17.
- A case study between Ericsson and Kiona in the Nordics showcased how connectivity and AI can be used to reduce energy consumption and carbon emissions of residential buildings by up to 15%12.
- In addition, efforts have been placed on reducing the carbon footprint during the construction process, by improving the efficiency of on-site operations as well as reducing waste:
- For example, HoloBuilder offers a 360-degree reality capturing system that uses AI to analyze construction sites, ensuring that materials are used efficiently and waste is minimized13.
- Similarly, AutoDesk Build provides construction productivity software, which uses AI to streamline planning and design, ensuring that resources are used efficiently14.
It is important as well to foster collaborative efforts across the construction industry in order to help develop new, more sustainable construction materials. One result of such collaboration is CarbonCure's technology, which introduces recycled CO₂ into fresh concrete, reducing its carbon footprint without compromising performance15. Once introduced into the mix, the CO₂ chemically converts into a mineral, creating a stronger concrete. The use of this technology has led to the reduction of over 340,000 metric tons of CO₂ emissions to date.
In short, there are several clear and impactful ways that AI has the power to rapidly help the construction industry reduce GHG emissions throughout the lifecycle of a building, from energy efficiency and material optimization to waste reduction and sustainable urban planning.
.jpg)
The retail sector, frequently overshadowed in discussions of environmental sustainability, is also undergoing a transformative shift, thanks to the integration of AI to not only help reduce retailers’ own carbon footprint, but also to reshape manufacturer and consumer behaviour towards more sustainable choices:
- One of the pioneering startups in this space is Refiberd, which is harnessing AI to tackle the colossal issue of textile waste16. They use AI to accurately detect fiber composition and the presence of contaminants in textile waste. Because recyclers are typically not able to process unsorted textile waste, their sorting system can enable recycling of up to 70% of textile waste that would otherwise be discarded.
- Vaayu's AI-driven Product Carbon Footprint (PCF) calculation model offers retailers insights into the carbon footprint of over 50 million fashion items18. Their collaboration with Klarna has equipped Klarna's user base of 150 million shoppers with a detailed breakdown of the environmental impact of their purchases. The goal of this partnership is to aid retail businesses in slashing their emissions and to encourage eco-friendly purchasing decisions.
- Similarly, Klarna has parnered with Clarity AI to present shoppers with AI-driven information on electronics manufacturers’ environmental efforts in the form of brand badges19. This provides a simple way for users to identify brands that are proactively addressing climate change, helping consumers shop more consciously.
Recent studies have delved into the psychology behind such sustainable consumer behaviors. By applying the stimulus-organism-response theory and the theory of planned behavior, researchers have found a "linkage effect" between online eco-friendly consumption habits and tangible offline sustainable actions. Furthermore, factors like passion and usability have been identified as indirect drivers that foster sustainable consumption, mediated by perceived value and customer loyalty20.
In conclusion, AI is proving to be an important tool in the realm of environmental sustainability within the retail sector. By helping to incentivize green consumer behaviour, AI can have a large positive impact on the overall supply chain ecosystem.
While AI's potential in addressing climate change is vast, its deployment requires careful consideration. The quality of data on which AI models are trained is paramount. Skewed or incomplete data can lead to incorrect predictions. This is why it is incredibly important for companies interested in utilizing AI to begin by ensuring high quality of historical data and putting strong data policies in place that will encourage good quality data going forward. In addition, when building any AI tool, including a feedback loop that allows human review and correction of errors will result in a positive cycle wherein model accuracy continues to improve over time.
Moreover, the "black box" nature of many AI models necessitates transparency to ensure trust and informed decision-making. It is important to build interfaces to AI model outputs so that the level of uncertainty in model predictions is clear to the end user, and ideally utilize methods that can allow the user to understand the driving factors that lead to any predictions made by an AI model.
Ethical considerations are equally vital. AI models must be designed with care to avoid unintended consequences. For instance, a model aimed at forest conservation must ensure it doesn't undervalue certain ecosystems or infringe upon indigenous rights. To avoid issues like this, it is important to consider potential bias in data used to train AI models, as well as implicit biases that may be inherent in the design of AI systems.
Finally, the environmental cost of running advanced AI models should not be ignored. Balancing the benefits of AI against its own carbon footprint is essential. As AI's role in climate action grows and the size of AI models increases, careful consideration of the environmental impact of AI itself will become increasingly important to ensure its responsible and effective use.
In conclusion, while AI has great promise to be a key driving force in our battle against climate change, its deployment is a delicate dance, one that requires nuance, understanding, and above all, a human touch.
At Yeji Data Lab, we pride ourselves on our deep and expansive AI expertise. With decades of experience under our belt, our seasoned team of AI experts has collaborated with partners across the transportation, construction, and retail sectors.
In the process, we've tackled many of these challenges and have consistently delivered innovative solutions. Our work includes projects with major ports, logistics companies, construction companies and retailers that implement tailored AI-powered solutions to transform their operational efficiency.
©Yeji 2023
