- Call: 0203 427 3507
- Email: innovation@clustre.net
While the world at large debates the future of artificial intelligence, the world’s logistics giants are quietly building proprietary AI empires that will reshape global commerce…
The most consequential artificial intelligence revolution isn’t happening in OpenAI’s boardrooms or Google’s research labs. It’s unfolding behind the closed doors of logistics companies, where billion-dollar stakes in proprietary AI systems are creating competitive moats that may prove insurmountable.
When UPS spent over a decade developing ORION – their core optimisation system comprising 1,000 pages of custom code – they weren’t just building software. They were constructing a fortress of intellectual property that no competitor could share or purchase. This represents a fundamental shift in AI adoption… because the most transformative applications are not bought; they are born from years of patient, secretive development.
Most AI discourse is rooted in the assumption that artificial intelligence is democratising technology… ChatGPT being an obvious and ubiquitous example.
But, in logistics, the opposite is proving true. The companies winning the AI race are not those who adopt the latest Chatbot variants. They are the ones building walls around their own innovations.
Consider Amazon’s 2012 acquisition of Kiva Systems for $775 million. This move wasn’t just about warehouse automation; it was also about denying competitors access to revolutionary robotics technology. Amazon didn’t just buy Kiva – they buried, resurrected and rebranded it as Amazon Robotics. Effectively, they ensured that no rival could access their IP. This calculated act of technological hoarding represents a new paradigm: AI as a strategic weapon rather than a shared tool.
This pattern repeats across the industry. Ocado’s proprietary robotic systems remain locked away from competitors. FedEx develops fleet management systems internally. Maersk and DHL guard their supply chain AI platforms with such secrecy that even their vendor relationships remain confidential. What we’re witnessing isn’t the democratisation of AI – it’s the creation of proprietary intelligence oligopolies.
Yet beneath this surface of technological secrecy lies a more nuanced reality. Even the most secretive logistics companies are building hybrid systems that combine proprietary innovation with strategic partnerships. But they are extremely selective about where to build and where to buy.
Walmart exemplifies this sophisticated approach. While developing proprietary demand forecasting models, they have fostered a deep partnership with Microsoft. One that’s designed to leverage Azure’s cloud infrastructure and OpenAI’s language models. Their five-year commitment to Azure represents more than shrewd vendor selection. It is a strategic bet that hybrid intelligence – combining internal innovation with best-in-class cloud capabilities – will outperform purely proprietary approaches.
Chris Chittock is one of the UK’s leading experts on streamlining logistics for SAP users. She is also the founder of Pivot – a member of our Clustre community and the ‘go-to’ solutions provider for many of the world’s biggest food and drink brands. Chittock is particularly impressed with Walmart’s broader technology strategy: “Walmart are supremely good at using tech to add value. Walmart Marketplace is a prime example. We are using it on more and more of our major integration projects for the world’s biggest CPGs. This powerful combination of AI and cloud could well challenge Amazon’s datacentre supremacy in the CPG space.”
This reflects a broader understanding among logistics leaders… namely that the most powerful AI systems aren’t monolithic solutions but orchestrated ecosystems. Walmart’s AI doesn’t just predict demand; it weaves together cloud computing, machine learning, and large language models to create something remarkable: artificial intuition about consumer behaviour.
When asked about the fundamental components of successful AI systems, Chittock identifies four critical building blocks: “What is the DNA of a successful AI system? I think there are four basic building blocks. One: Great Data (both internal and external third-party). Two: Context. Three: Applications. Four: Cloud. What’s more, if companies have SAP, they have a built-in head start. SAP is the most data hungry application in the world, because the applications touch every business process and provide the context that so many companies are coding from scratch. AI is nothing without great data and context.”
The most compelling aspect of AI-driven logistics is to be found not in its computational power but in its capacity for unexpected discovery. When Hovis implemented AI, they expected better demand forecasting. What they discovered was something much more profound: a hidden relationship between weather patterns and the human appetite for bread.
This wasn’t programmed into the system. Interestingly, it was another Clustre firm that helped Hovis to make this discovery. Using AI, they identified correlations that human analysts had never previously considered. This led directly to radically improved forecasting and a dramatic 50% cut in wasteful over-production. But the real value wasn’t the waste reduction. It was the revelation that atmospheric pressure, temperature variations, and precipitation patterns influence consumer behaviour in ways we’re only beginning to understand.
This exemplifies AI’s most transformative capability: pattern recognition that transcends human cognitive limitations. These systems don’t just optimise existing processes; they uncover hidden relationships that redefine our understanding of consumer behaviours, supply dynamics, and market forces.
Tesla may capture headlines with attention-grabbing autonomous vehicle demonstrations, but the real breakthrough in logistics is happening elsewhere. Companies, such as Waymo, are pioneering methodical development programs and, in the process, they reveal the sophistication required for truly transformative AI.
Waymo’s system combines Google’s Gemini multimodal language model with their own, proprietary, Waymo Foundation Model – a hybrid intelligence system trained on millions of miles of real-world driving data. This isn’t just autonomous driving; it’s the creation of artificial experience learning. Waymo’s systems don’t simply follow rules – they develop an intuitive understanding of complex, unpredictable environments.
The combined architecture really can be greater than the sum of the parts. Proprietary sensors, custom hardware and internally developed vision systems – when combined with world-class cloud AI capabilities – can deliver significantly enhanced value. Waymo are not just building self-driving car. They are creating a template for proprietary innovation and strategic partnerships to deliver AI capabilities that far exceed their individual merits.
The most sophisticated logistics companies have recognised a crucial truth…
While AI algorithms may be proprietary, the underlying infrastructure increasingly demands cloud-scale capabilities. This has led to strategic partnerships that blur traditional vendor-customer relationships.
Microsoft Azure, Google Cloud AI, and AWS are not only providing computational resources, they are also integral to proprietary AI development. Companies now leverage cloud platforms for machine learning, model training, data processing and, increasingly, for accessing cutting-edge capabilities… such as large language models that would be prohibitively expensive to develop internally.
Many logistics companies also maintain competitive differentiation through proprietary algorithms and data while leveraging shared cloud infrastructure for scalability and advanced capabilities. The result is AI systems that combine the best of both worlds… competitive advantage built on world-class foundational technology.
Perhaps the most under-appreciated aspect of logistics AI is the optimisation of data that established companies already possess. UPS’s ORION system isn’t just sophisticated code – it’s trained on decades of delivery data, customer patterns, and operational experience that no competitor can replicate.
This data advantage compounds over time. Every package delivered, every route optimised, and every customer interaction feeds back into the AI system, making it progressively more sophisticated. New entrants don’t just face technological barriers; they confront an insurmountable experience gap.
Amazon’s logistics AI benefits from insights drawn from billions of customer transactions, millions of delivery routes, and vast warehouse operations. This operational data creates training datasets that no external AI system, however sophisticated, can match. The companies with the most data aren’t just building better AI – they are creating self-reinforcing competitive advantages.
What emerges from this landscape is a sobering insight:
The key imperative is to identify where proprietary advantage matters most and leverage partnerships for everything else.
The logistics industry stands at a true tipping point. Artificial intelligence has evolved beyond experimental curiosity to become the foundation of competitive advantage. Companies that recognise this shift and invest strategically in AI – whether proprietary, partnered, or hybrid – will define the industry’s future.
This transformation extends beyond operational efficiency to fundamental business model innovation. AI-powered logistics companies aren’t just moving goods more efficiently; they are reimagining the relationship between supply and demand… prediction and fulfilment… efficiency and resilience.
This ‘arms race’ has profound implications…
The winners won’t be those who adopt AI fastest, but those who understand where to build, where to buy, and where to partner. Success demands not just technological sophistication but strategic wisdom about how to construct sustainable competitive advantages in an AI-driven world.
The industrial revolution mechanised physical labour. The digital revolution computerised information processing. The AI revolution is automating intelligence itself. The companies that master this transition will reshape global commerce for decades to come.
Ian Spencer is a founding partner of Clustre, The Solution Brokers.
It is no accident that AI is an area of special strength within Clustre. Wherever you are on your AI journey… whatever your expectations and destinations… Clustre has the greatest concentration of world-class AI skills to help you.
Let’s talk – openly and without fee. Call Robert Baldock – Clustre’s MD – on 07768 402131. robert.baldock@clustre.net
