Analyst Insight: ESG has become a critical business focus in recent years, and supply chains are arguably the most crucial component of sustainable business practices. With the advancement of AI technology, we are increasingly seeing companies turn to the technology to support supply chain sustainability, but the environmental benefits may be overridden by the energy needed to power the technology.

AI has the potential capabilities to optimize logistics, reduce waste and improve energy efficiency within supply chains by leveraging advanced analytics and real-time data. All these factors can improve sustainability, and move supply chain operations towards achieving ESG goals.

The technology appears to have taken various business sectors by storm, and the supply chain and logistics sector is no different. According to a survey of 3PLs last year, 84% of providers describe AI and machine learning as among the technologies most likely to have industry-wide impacts.

However, businesses should not go into implementation blindly.

Making Sustainability Gains

By analyzing historical data, and using machine learning algorithms, AI can optimize planning and transport operations, thereby reducing carbon emissions. One of the key areas where AI can be impactful is route optimization. AI algorithms can analyze various factors such as traffic patterns, weather conditions, road conditions and delivery schedules to determine the most fuel-efficient routes for transport vehicles. By minimizing unnecessary detours and lengthy routes, fuel consumption can be optimized to reduce carbon emissions.

The technology can also facilitate manufacturing efficiencies through improved equipment utilization, reduced downtime, and more accurately predict demand. This in turn can optimize inventory levels, reducing excess stock and wastage. It also removes the need for rush deliveries that lead to unnecessary transport emissions.

Acknowledging AI’s drawbacks

While AI has great potential to drive improvements in sustainability, it is not without its challenges. AI algorithms require large amounts of high-quality data to operate. This can be hard to obtain and manage within global supply chains. By nature, supply chains rely on various tech systems such as inventory management, finance and transport systems. Streamlining the data across these systems can be challenging if they aren’t directly integrated. Siloed data systems can’t provide AI systems with the necessary support. as they need to be regularly fed robust data sets. Therefore, without greater data management across the supply chain, AI can’t be used to its maximum potential.

AI’s energy consumption levels are also a concern. This little-recognized issue was raised in a paper by Alex de Vries, a researcher at the Vrije Universiteit Amsterdam. He claimed that if every Google search became an AI interaction, the electricity needed to power this could amount to the same consumption level as the country of Ireland, at 29.3 TWh per year. Considering the quick growth rate of the technology, there is a chance that any environmental benefits may be overridden by the energy needed to power the technology, at least in its current form. Therefore, a complete switch to the technology is not currently environmentally responsible.

Taking a Practical Approach

With this in mind, supply chains need to take a balanced approach to implementing AI. Businesses need to enhance sustainability while also considering data requirements and energy consumption. Businesses should be encouraged to focus on the areas of the supply chain where AI can have the greatest impact. Transportation is an aspect that could significantly benefit from AI support, as it contributes around 24% of global CO2 emissions. Therefore, implementing AI-enabled route optimization is an appropriate use of the technology for greater sustainability, without being too energy intensive.

It is also important to implement continuous monitoring of the AI systems to ensure that any inefficiencies and opportunities for improvement are identified. Businesses should regularly assess the energy consumption and the accuracy of the system to maximize its return on investment.