Supply chain sustainability is no longer about greener infrastructure alone; it is about enabling growth without proportionately increasing resource consumption. Renewable energy, electric mobility, efficient facilities and automation help, but smarter operational decisions matter equally. Poor scheduling can waste energy, underutilised vehicles consume resources and excess inventory occupies space unnecessarily.
Sustainability therefore depends on better decisions around inventory, capacity, routing and equipment. Capgemini’s 2025 research found 82% of organisations plan to increase sustainability investments, while 67% cite profitability and efficiency as drivers. Yet only 21% have detailed transition plans, making execution the critical challenge.
Utilisation is becoming a sustainability metric
Supply chains have traditionally met rising demand by adding capacity, more warehouses, vehicles and equipment. But with growing pressure on land, energy, materials and capital, the focus must shift towards using existing capacity more productively. Freight logistics accounts for around 8% of global emissions, while AI could potentially reduce logistics emissions by up to 15% through better operations, capacity utilisation and modal choices. The sustainability equation is therefore not only about cleaner fuels or energy sources; it is also about how efficiently assets are used. Underutilised trucks, unnecessary journeys, inefficient warehouse routing and idle equipment all consume resources without delivering equivalent value. The priority is to generate more output from every unit of energy, space and capacity.
Automation needs system-level intelligence
Automation has already transformed warehouse operations, from robotic movement and automated storage to software-driven coordination. But as facilities become more complex, optimising individual machines is no longer enough. A warehouse with hundreds of automated assets can still face congestion, unnecessary movement and uneven utilisation.
Orchestration addresses these challenges by coordinating assets across the network reallocating tasks, positioning inventory based on demand and scheduling equipment around operational needs. The goal is to reduce idle time and unnecessary movement while maximising existing capacity. McKinsey estimates AI can reduce inventory by 20–30% and logistics costs by 5–20% through forecasting and optimisation.
Data must move from reporting to action
Supply chains already generate vast amounts of data through warehouse systems, fleet platforms, sensors and control towers. The challenge is turning this information into timely operational decisions. Energy systems can flag unusual consumption, but greater value comes from identifying causes and adjusting equipment schedules or workloads. Fleet data can highlight underutilised vehicles, enabling load consolidation or return movements.
Warehouse data can identify congestion before it affects throughput. This requires connecting insights across operations, linking energy with throughput, inventory with demand and vehicle utilisation with shipment patterns.
Inventory belongs in the sustainability equation
Inventory is typically viewed through working capital and service levels, but its environmental cost is equally significant. Every unnecessary unit consumes raw materials, manufacturing capacity, packaging, transportation, storage and handling resources. If it becomes obsolete, those resources deliver little economic value. Better forecasting can reduce this waste.
McKinsey estimates AI-enabled distribution applications can lower inventory by 20–30% through improved forecasting and optimisation. The goal is not to minimise inventory indiscriminately, but to improve decisions on what to produce, where to position it and when to move it. Better decisions reduce capacity held against uncertainty, lowering costs, space requirements and resource consumption.
Intelligence also carries an environmental cost
AI’s sustainability case must account for the resources required to run it. The IEA projects data-centre electricity consumption will more than double to around 945 TWh by 2030, driven partly by AI. For supply chains, AI adoption therefore needs a measurable efficiency case: reducing unnecessary movement, excess inventory or underutilised assets. Technology creates sustainable value only when the resources it consumes are outweighed by the resources it helps save.
Sustainability needs to enter the operating model
Supply chain sustainability needs to move beyond broad metrics to operational measures such as energy per order, throughput per square metre, carbon per tonne-kilometre and energy per automated movement. These metrics connect resource efficiency with cost, service and productivity.
The advantage will not come from building the largest warehouse, fleet or most automated facility. It will come from extracting more value from every asset already deployed.
The future supply chain must build what is necessary while making every machine, kilometre, square metre and unit of energy work harder. Sustainable growth will depend not on adding capacity indefinitely, but on delivering more with the resources already available.
