After years of headlines and high expectations, artificial intelligence in supply chain operations has reached a turning point. The conversation has shifted from potential to proof. At the recent Retail Supply Chain Summit in Sydney, I had the opportunity to host a roundtable with several of Australia’s leading retailers. Our discussions focused on what’s really working with AI in supply chains today – and what’s not – as organisations move from experimentation to practical, value-driven deployment.
When AI works, and when it doesn’t
The discussion revealed a mix of success stories and growing pains. One retailer had implemented robotics automation in its fulfilment process but ultimately found that order volumes were insufficient to justify the investment. Eventually, the retailer reverted to manual processes. Underscoring a key lesson: automation must align with business scale and order profiles to deliver meaningful returns.
Other participants were exploring AI’s potential in demand forecasting and marketing personalisation, drawing on both publicly available information and customer data to better anticipate demand and tailor promotions. These initiatives showed promise but also highlighted the need for robust, accurate data as the foundation for any AI model.
Read also: Retail Fulfilment Show 2026: Redefining the future of retail logistics
In another case, a business experimented with drone-based stocktaking using RFID and image recognition technology. Pallet wrap reflecting in warehouse lighting made barcodes difficult to read, illustrating the iterative nature of AI integration. Across these examples, one clear message emerged: AI adoption is not a single leap forward but a continuous cycle of testing, refinement and improvement.
From the warehouse floor to the supply chain network
AI is already creating tangible results for Infios customers across several areas. In planning and forecasting, machine learning models are improving demand predictions around seasonal peaks, promotions, and unexpected surges. In warehouse operations, AI-driven slotting optimisation is enabling continuous analysis of SKU movement, automatically reconfiguring warehouse layouts to boost picking efficiency.
Digital twin technology is also gaining traction. By virtually simulating warehouse layouts or network design changes, businesses can evaluate multiple operational scenarios before committing to physical adjustments. This approach reduces both risk and cost while accelerating optimisation.
The most compelling results are emerging where AI orchestration connects directly with automation. By linking AI-based order orchestration, labour planning, and slotting optimisation with autonomous guided vehicles, companies can achieve higher levels of coordination and throughput. This fusion of predictive intelligence and physical automation represents one of the most promising frontiers in supply chain execution.
Read also: Nearshoring: The myth of distance—and what really matters
The next frontier: predictive and accessible intelligence
At Infios, the focus is on building an Intelligent Supply Chain Execution suite that embeds machine learning across the entire platform. The goal is to move beyond task automation toward predictive orchestration – enabling systems to anticipate disruptions, model alternatives, and recommend proactive actions.
As supply chain networks grow more complex, decision-makers increasingly want visibility into external factors such as weather patterns, transport delays, and port congestion. Predictive AI can help them respond dynamically, identifying when to reroute shipments or adjust transport modes before issues escalate.
Another key development area is conversational AI, designed to make advanced insights more accessible to non-technical users. Instead of relying on IT teams to extract reports, warehouse and transport managers will be able to interact with systems in natural language, asking direct questions and receiving actionable answers in real time.
From buzzword to business driver
The roundtable made one thing clear: AI has moved beyond the stage of being a buzzword. It is now a proven enabler of operational efficiency and resilience – but success depends on getting the fundamentals right.
Clean, reliable data, a skilled and informed workforce, and a commitment to iterative learning form the foundation of any successful AI strategy. The companies willing to experiment, measure results, and adapt their approach will be the ones leading the next phase of digital transformation. The industry doesn’t need more speculation about
what AI might do. It needs clear, evidence-based examples of where it delivers measurable value – and where it doesn’t. The path forward lies in moving beyond hype and focusing on practical, data-driven impact.
