Rerouting Supply Chain Inefficiencies With AI

Forbes

a day ago

Clemente Theotokis is the Cofounder of Zeus, a London-based startup transforming logistics with data, AI and automation.
The logistics industry is no stranger to complexity. It requires a delicate balancing act and urgency, with little room for error.
Today's volatile supply chain environment only adds to the challenge, especially with customers who want next-day deliveries that strain supply chains. As profit margins shrink, the pressure to do more with less intensifies. Frequent disruptions are exposing deep inefficiencies in supply chains, reinforcing the urgent need to rethink and fine-tune every link in the chain to ensure long-term resilience and profitability.
One of the most persistent and costly inefficiencies in freight logistics is waste, particularly when trucks aren't filled to capacity. This leads to higher fuel consumption, increased carbon emissions, and drivers left waiting idle for their next load.
Estimates suggest that 20%-35% of truck miles worldwide are driven empty, and even "full" trucks often carry less than capacity. While pinpointing a definite number remains difficult in such a fragmented sector, one thing is clear: progress has been remarkably slow. You can often see it yourself: A raised fourth set of wheels on trucks signals they're running empty or unloaded.
As AI begins to reshape the traditional operating models relied upon by the logistics industry, data-driven transformation is propelling those companies that are planning ahead. Hesitations around adopting new tech remain, including concerns about implementation costs, limited in-house expertise, poor data quality and a growing skills gap in an aging logistics workforce. Without the right training, even the most advanced tools risk being underused or misapplied.
I have seen firsthand that the real challenge is reshaping how teams think and operate. The most meaningful gains come when companies commit to a deeper cultural shift, and those that succeed will be the ones willing to challenge legacy thinking, upskill their teams and embrace change as a continuous advantage. AI gives us the power to see and solve problems faster, but it will be leadership that sets the pace.
According to a report by McKinsey, AI has the potential to generate $2 trillion in value across the supply chain and manufacturing sectors by the end of 2025 via smarter, more efficient logistics. Real success hinges on the effective use of the tools at our disposal. It's no use having all the gear and no idea. So, where to begin?
Start With Smarter Data
First, identify the inefficiencies and gaps AI can help address. Pinpointing these specific issues early lays the groundwork for a clear and effective AI implementation strategy. But you need to dig deeper into your data to really understand what's going on across your supply chain.
AI is only as good as the data it's fed, so start with a thorough audit, honing in on key areas where things tend to go off track. For example, in freight management, are delays common? Are spot-market rates overused? In the warehouse, how often do stockouts or overstocking occur? Are fuel and driver costs high due to poor route optimization? From a customer's perspective, are shipments often misplaced or delayed, or lacking real-time updates?
If your datasets are messy or incomplete, predictions will be inaccurate and unreliable. Therefore, it's imperative to gather, clean and standardize data across your operation.
Start with the fundamentals: shipment records from your TMS, carrier tracking updates, inventory levels and demand signals. By integrating data from transport, warehousing and fulfilment systems via APIs, create a unified view of operations. Once gathered, this data should be cleaned and standardized to remove inconsistencies and fill gaps. This helps ensure that performance metrics such as delivery time, cost per pallet and stock accuracy are reliable. With a solid data foundation, external factors like fuel prices or carrier availability can be incorporated into planning with greater confidence.
As for fragmented data, using what's already available is a great place to start. Simulation and modelling help test and validate outcomes early, well before any widespread deployment or commercial commitment. It's a low-risk way to explore potential benefits and overcome early hesitation by reducing the fear of getting it wrong or wasting resources. You can't run before you learn to walk, so it's about making better use of what's already to hand.
Turn Data Into Action
Once a solid data foundation is in place, AI can start delivering real impact. Retailers, for example, can use AI-driven demand forecasting to better predict and manage spikes during seasonal promotions to avoid both stockouts and excess inventory. Haulers can optimise routes in real time to avoid traffic delays and cut fuel costs. Meanwhile, manufacturers can anticipate possible risks such as supply chain disruptions caused by severe weather or political events like new tariffs, and adjust operations accordingly to keep things moving smoothly.
Consolidating operational data through AI enables real-time analysis across the entire supply network. Unlike traditional models that rely on fixed rule sets, simulation-based approaches can capture real-world complexity across multiple layers of the supply chain, such as demand and inventory shifts, route optimization and clustering fragmented deliveries into full-load journeys. By comparing current logistics performance with AI-generated scenarios, it's possible to identify opportunities for improvement such as reducing unused fleet capacity, resolving disconnected orders or streamlining network flows. These simulations can improve forecast accuracy, resource utilisation, and overall cost efficiency, often delivering double-digit percentage uplifts in performance.
As AI matures, its power to transform logistics will only grow stronger. While it may not offer the perfect quick fix, it represents a long-term investment. Naturally, there will be bumps along the way, but a huge part of navigating those challenges is education. Companies are learning what the technology can do, why it matters, and how to put it to work in the correct way. While the industry is undeniably complex, with the right knowledge and tools, complexity can become a real competitive advantage, making way for a smarter, more efficient future.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?