THE NEXT paradigm shift in the leadership of logistics operations will spring from using AI to drive both logistics operating model design and day-to-day management. Managers must find a path for their business or risk becoming uncompetitive in their logistics operations.

My teams and I have spent 40 years – at companies such as Marks & Spencer, Fat Face, Just Group (AUS) and Debenhams  – improving the design and delivery of operating models, primarily in retail, but also in the airline and food sectors. Managing the flow of goods from supplier to customers and fulfilling demand across global channels, we have made careers in delivering change in logistics. I have written business cases in excess of £1 billion in value. I’m not unusual and this reflects how far we still need to travel to achieve optimised supply chain and logistics operations.

Over these decades many things have changed. Many developments have reduced costs and improved customer service but principles of good practice have not changed and remain elusive for many organisations. Our work as leaders is not complete.

We have seen developments in automation and supply chain planning while multi-channel retailing has made those improvements essential. Automation and new handling techniques have reduced costs as a percentage of sales but they have also constrained the flexibility of some assets.

Same old, same old

Two core concepts still appear to undermine operational performance:

1. The imbalance between supply delivery and demand fulfilment – creating operational costs and customer service risks.
2. Poor efficiency in the core elements of the operation across freight, warehousing, transport and fulfilment due to poor practice and visibility.

Costs and service have improved – but in the context of the same old issues, such as:

• Excessive stock and poor SKU focus.
• Poor planning, throughout the year and in particular for Peak.
• Underutilisation of key logistics assets such as warehousing, automation and equipment.
• Labour and automation readiness compared to the ‘new’ trading plan.
• Change complexity in logistics is a corporate not functional risk.

The next evolution of our industry will be driven by AI but its effectiveness can only be delivered by leaders in supply chain and logistics – YOU. So, what do you need to give your business the competitive advantage from the evolving AI opportunity?

AI – the aspiration

The AI opportunity is for us as leaders to drive a new level of understanding, operating model options and new levels of efficiency, and ultimately a greater opportunity for SC&L to be a key corporate competitive advantage.

The opportunity derived from AI and large data sets may provide us with more solutions and options – in minutes – than our teams could create and evaluate in months of work.

If AI could….

1. Show the optimised model for my supply and demand. To do this it needs to be able to recognise my constraints – current model and contracts; and show the value creation from the change of each individual operational parameter.
2. Drive change in how I manage my logistics partners. Support the changing of the relationship – operational and commercial – between the players in the model. What are the best commercial structures for my optimised operating model?
3. Improve the understanding of operating model through visualisation. Many functions impact the Op Model performance. AI with its speed and visualisation capabilities should be the next step change in decision making – not just in logistics – but corporately. A suitable solution will help buying, merchandising, and all trade departments make improved commercial decisions with far greater awareness of operating model impact. A key step change in the culture and ways of working.

This is an enormous challenge and, as I see it, this stage has not been delivered. The challenge is not a quick one and the path to delivery is paved with potholes.

Current tech players

The current market is myriad and complex. It will therefore evolve and the winners will be flexible with their IT architectures to take advantage of evolving capabilities.

Gartner outlines the key players at the moment as Siemens, IBM, Microsoft and specialists AnyLogistix, Coupa and Dassault. However, if you look the components of an AI solution, it also takes you into the realm of API tools, simulation engines, optimisation engines and the tools to digest structured and unstructured data. Tools can be prone to hallucinations. Beware of technology firms not exposing their component capabilities but promising the world.

The simple analogy for a car enthusiast is there is no point buying a great engine, with square wheels, and only two seats for your family of five. It takes a full set of components that align to keep the family safe and happy.

As I explore the component technologies required to serve the leadership of a supply chain and logistics operation, we must present our requirements to the technology firms as we have to deliver the daily operation to our customers.

The risk leaders need to manage

Technology has been the greatest friend and most fiendish foe in my career. It has been the driver of many a step change. Likewise, it has undermined many an operation. AI will be no different – so learn from the past technology revolutions that all required operational leadership.

• Automation – required careful specification, maintenance and management of product specification and velocity.
• IT architectures and stock pots – stock inaccuracy across the architecture left us confused between physical, system and the control of stock assets and customer service.
• Planning and forecasting – new levels of planning demanded greater responsiveness from the operation.

From the lessons of the past, we can manage the future. We as leaders need to play our role in each CSF.

AI critical success factors

IT architecture: an IT leadership challenge but drives issues into the operation day-to-day. We have a voice to contribute to ensuring architecture and the new level of AI operational suggestions does not undermine smooth execution. The three core layers of architecture all require SC&L leadership input.

1. Data warehouse – can you get aligned data to drive feasible options?
2. Computational layer – make sure the calculations are verified.
3. Visualisation – it needs to work for all teams, cross functionally and with partners.

Tools: The current debate about Claude’s Fable is noteworthy (withdrawn and tweaked due to security concerns); what will be tomorrow’s best AI tool for logistics? You need to be able to plug and play with multiple tools as it is unproven which AI tools work best in logistics.

Data: It’s the same old story – rubbish in rubbish out – but exponentially now even more important. You need to ask what level of structured and unstructured data accuracy the AI is working with. Also, what are the assumptions fed into the system against which it calculates?

Performance: These are big data models, and as we know slow systems often equate to slow inaccurate operations. If we are to drive our operations from AI data it needs to be quick!

Verification: Some of what is presented by AI tools is brilliant. Some of it is not. Blindly assuming the former could lead to disappointment. In design and testing verify calculations and appropriateness for your operation.

Guard rails: In my work, I am designing an optimised unconstrained solution and then putting on guard rails and assumptions relevant to my business into the system. The system then produces a constrained optimisation so I can evaluate the impact of each of my constrained parameters. We do not all live in a green field.

In summary, my advice is:

1. Technology – choose the tools knowing their strengths and weaknesses. To do this you need a knowledgeable IT dept/colleague.
2. Know your business – understand the value chain and the gap to unconstrained optimisation and your realistic constrained optimised solution. What value and improvement are you pursuing?
3. Partner up – suppliers, third party logistics providers and technology firms can have input into the design of the solution for operators and for customers.
4. Experiment, experiment, experiment – no one won the war by trying it first in battle.
5. Accept it’s a journey – so start now – and accept the route is unclear.

I am collaborating with selected technology and consultancy partners to build a solution that meets the requirements of a leader in operations. I am also pulling groups of us together to share our experiences and views of the right route forward.

I welcome input and look forward to your views. I can be contacted on [email protected] or 07766 504240.