Rohit Laila is a veteran of the logistics world, a man who has spent decades watching the gears of global supply chains turn from the inside. With a career that spans the heavy lifting of physical delivery and the high-level orchestration of digital supply chain planning, he has become a leading voice on how technology actually hits the warehouse floor. In an era where every boardroom is buzzing with the promise of artificial intelligence, Laila brings a grounded, practical perspective to the conversation. He understands that while a company might have the capital to buy the latest software, the real challenge lies in the “connective tissue” of data and human trust that makes those systems work.
In this discussion, we explore the stark reality of modern logistics automation, where massive investments often struggle to produce autonomous results. We dive into the critical differences between routine tasks and high-stakes strategic decisions, the evolving role of the human planner from a data entry clerk to a cross-functional leader, and the cultural hurdles that keep most organizations from fully handing over the reins to AI.
Many organizations have invested between $3 million and $10 million in supply chain planning automation, yet true autonomy remains rare. What are the specific technical and data foundations that companies often overlook during these expensive rollouts, and how can they bridge the gap between high spending and AI readiness?
It is a sobering reality that 51% of senior leaders have funneled between $3 million and $10 million into these systems only to find that their AI is essentially idling in the garage. The most common oversight is the “data silo” problem, where information exists but isn’t clean or connected enough for a machine to make a reliable decision. I have seen companies spend millions on shiny interfaces while their underlying data quality remains so poor that the AI generates “hallucinations” rather than actionable insights. To bridge this gap, leaders must stop viewing AI as a plug-and-play purchase and start treating it as an ecosystem that requires a foundation of high-quality, real-time data and a workforce trained to interpret it. True AI readiness is about the maturity of your digital infrastructure, not just the size of the check you write to a software vendor.
While routine tasks like order prioritization are prime candidates for automation, high-stakes choices regarding facility placement and inventory strategy still require human judgment. How should leaders categorize decisions based on risk and complexity, and what specific criteria determine if a process should be fully automated or merely augmented?
The distinction usually comes down to the “weight” of the consequence and the shelf-life of the decision. Routine choices like daily replenishment or order prioritization are high-frequency and low-risk, making them perfect for a machine that can process thousands of variables in seconds. However, when you are talking about where to build a $50 million distribution center or how to hedge against a global trade shift, you need the nuanced, emotional, and historical context that only a human brain provides. Leaders should categorize decisions by weighing four key factors: value, complexity, risk, and the specific need for human intuition. If a mistake in the process could bankrupt a product line or sever a key supplier relationship, that is a clear signal that the AI should augment the human, not replace them.
As technology begins to handle more administrative data tasks, the role of the human planner is shifting toward cross-functional coordination. What new skill sets must planners develop to effectively manage these digital systems, and what does a day-to-day workflow look like when a human transitions from a data entry role to a strategic decision-maker?
We are moving away from the era of the “Excel jockey” who spends eight hours a day cleaning spreadsheets and moving into the era of the “Supply Chain Architect.” Planners now need to master the art of cross-functional coordination, learning how to negotiate between sales, finance, and logistics to align on a single version of the truth. A typical day no longer starts with data entry; it starts with reviewing AI-generated scenarios, where the planner evaluates different outcomes and makes the final call on which path to take. This shift requires a deep understanding of how to set the “rules of engagement” for the machine, essentially transitioning from a worker who does the task to a manager who supervises the digital system. It’s a move from the tactical weeds to the strategic mountain top, and it requires a much higher level of emotional intelligence and communication.
Software deployment is often used as a metric for success, but manual workarounds frequently persist behind the scenes. What specific metrics should a company track to prove that an AI tool is actually improving decision quality, and how can they ensure staff are not bypassing the technology in favor of legacy methods?
You have to look past the “go-live” date and start measuring what actually changes in the daily grind of the planning office. I always tell my clients to track the decline of manual workarounds—those secret spreadsheets that planners keep under their desks like a safety blanket. If your team is still spending hours “adjusting” what the AI tells them, the system has failed to earn their trust or the data is wrong. We should be measuring decision quality by looking at the accuracy of the automated outputs versus the manual ones and tracking whether the speed of those decisions actually improves. If the staff feels the AI is a burden rather than a tool, they will bypass it every single time, so adoption rates and user sentiment are just as important as ROI numbers.
Forecasts suggest that by 2030, only 5% of organizations will allow technology to make even 10% of their planning decisions autonomously. What are the primary cultural and organizational barriers preventing a faster transition, and what step-by-step approach can a company take to gradually increase the “decision-making authority” of their AI systems?
The primary barrier is a fundamental lack of trust in the “black box” of AI, fueled by a fear that a machine might make a catastrophic error that a human would have spotted. Even though companies are spending millions, there is a deep-seated cultural resistance to letting go of the steering wheel, especially when 83% of organizations are still grappling with the basics of automation. To overcome this, I recommend a tiered progression: start by letting planners simply assess AI output, then move to a collaborative model where the human and machine solve problems together. Eventually, once the system has proven its reliability over hundreds of cycles, you can move to the final stage of “exception-based planning,” where the machine acts on its own and only alerts the human when something goes wrong. It’s a marathon of trust-building, not a sprint of technical installation.
What is your forecast for the future of autonomous supply chain planning?
My forecast is that we will see a slow, deliberate climb where the “trust gap” finally begins to close as systems become more transparent. While the 2030 prediction of 5% autonomy seems conservative, it reflects the reality that supply chains are incredibly volatile and human intuition remains our best defense against the unexpected. By 2030, the organizations that win won’t necessarily be the ones with the most autonomous machines, but the ones who have best integrated AI to handle the “noise” of routine data so their humans can focus on the “signal” of strategy. We are heading toward a future of “human-centric autonomy,” where the machine does the heavy lifting, but the human heart and mind still hold the final veto.
