Aptean’s Ben Glossop tells Peter MacLeod that he believes artificial intelligence will transform logistics through incremental improvements rather than overnight revolution.
AI may dominate conversations about the future of logistics, but for many operators the immediate question is what can it do for my business today? For Ben Glossop, Aptean’s Vice President of Sales, the answer begins not with futuristic autonomous fleets, but with mundane tasks that currently consume hours upon hours of employees’ working days. He says processing emails, updating orders, responding to delivery-window changes, and checking whether planned schedules reflect actual performance are all strong candidates for automation.
Glossop, whose background encompasses extensive experience with transport routing specialist Paragon, believes the industry should resist portraying AI as an overnight revolution, but more of an evolution in which software assumes routine responsibilities, allowing people to concentrate on decisions requiring experience and judgement.
Automating the everyday
One immediate opportunity lies in managing the constant flow of small changes that characterises transport operations. A customer might request a different delivery time or amend an order, and whilst each request may be straightforward, handling hundreds of them creates a substantial administrative burden and introduces margin for errors.
Glossop describes work with a customer whose inbox receives between 70 and 100 emails a day requesting changes to delivery time windows. Here, an AI agent can extract the request, update the schedule and assess its operational consequences. Where an amendment falls within agreed tolerances for additional mileage and time, the system can accept it and respond automatically. If the change would require another vehicle or introduce significant cost, the request is referred to a person for a decision.
AI is not given unrestricted authority to alter a transport plan, as the customer sets the parameters within which the agent can operate. Greater autonomy can then be introduced as confidence develops. “Do we automate all of them straight away? No,” says Glossop. The system can ask for approval repeatedly, learn from decisions and gradually take on more responsibility for familiar tasks. Human judgement remains essential when requests fall outside agreed parameters. Glossop argues that industry-specific – or vertical – AI has an advantage over generic tools because it learns from customer decisions and understands logistics terminology and constraints.
Existing Data
Aptean is also working to make its transport planning software easier to use. Its Ask Paragon functionality allows users to interact with the system in natural language when interrogating it about route performance or the potential effects of changing delivery windows. Previously, extracting the full value from routing software would require specialist knowledge of advanced settings. But this more conversational interaction lets planners explore improvements without manually reconfiguring multiple parameters.
The wider ambition is to connect information across business applications. Through Aptean’s AppCentral platform, Glossop envisages users querying data across systems rather than treating TMS, WMS and other applications as isolated environments.
The concept is effectively an orchestration layer above individual systems. An operator could identify a problem at a higher level, with agents making adjustments in relevant applications. This could help expose bottlenecks and show how decisions affect the wider operation, depending on system integration and data quality.
Proving AI’s Value
For all the enthusiasm surrounding AI, the business case always has to stand up to scrutiny. After all, what’s the point in AI just for AI’s sake? When I asked Glossop whether Aptean’s AI applications have demonstrated measurable improvements in transport efficiency, labour productivity or emissions, he is candid about the current stage of development. “The main metric at the moment is time saved,” he says. The initial objective is to automate time-consuming processes, or checks that are neglected because employees lack capacity.
But the opportunity extends beyond administrative savings. Glossop describes an agent being developed to identify under-utilised routes and contact customers who have not placed their usual orders. If, for example, a regular route is forecast to operate below 80% utilisation the following day, the agent could prompt customers to order, then alert the account manager if necessary.
Additional orders could improve vehicle utilisation, measured through load factors and the proportion of delivery capacity used. Some benefits arise because a task is completed more consistently rather than faster.
Glossop cites an implementation in which operational staff initially claimed that routing software had increased scheduling costs by 20%. The apparent deterioration resulted from a mismatch between old and new methods. Previously, planners had considerable freedom and did not consistently adhere to delivery windows, driver duty times and vehicle restrictions. The software, by contrast, was configured to respect hundreds of operational constraints, and therefore the approaches were not comparable: the new system had to meet constraints that had not always been enforced, as well as agreed service commitments. Technology therefore cannot be judged on headline costs alone; service performance and compliance matter too.
The Human Side
Even when technology works as intended, implementation can flounder if the people buying the solution and those operating it are not aligned. Senior management may focus on savings and return on investment, while employees worry about how automation will affect their roles. “If we walk into a company and just say, ‘Do as you’re told, because this is the right way of doing it’, there’s a lot to take on board,” says Glossop.
I suggest that winning the ‘hearts and minds’ of those charged with working with the technology is essential, and Glossop agrees. Successful adoption requires employees to understand what the technology is intended to achieve and to help shape how it works, and Aptean’s implementation process takes this into account. Those closest to a process often know where inefficiencies lie, but may be reluctant to share that knowledge if they fear automation will make their jobs redundant.
Glossop has seen the role of the transport planner evolve over the years, primarily in the shape of a move towards more strategic work, but still acknowledges that implementation teams need technical expertise and an understanding of the pressures on employees who must keep deliveries moving while systems change.
Delivering Results
The next stage of Aptean’s approach is reflected in Aptean Ascent, a programme intended to combine AI, industry expertise, process changes, and ongoing support to achieve defined business outcomes.
Rather than simply supplying software and leaving customers to realise its potential, the model involves working with the business to identify opportunities, implement solutions and assess results. Aptean’s published description of Ascent emphasises benchmarking performance, agreeing measurable objectives and validating outcomes against those baselines.
Glossop describes the early-stage thinking as potentially resembling a managed service, with Aptean taking greater responsibility for the technology and operational changes needed to achieve agreed results. He also discusses possible revenue sharing, with fees linked to savings delivered.
The approach addresses a tension in software purchasing, namely that customers want savings guaranteed while results depend partly on the decisions they control. Operators must agree a balance between vehicle utilisation and service flexibility, and then establish how success will be measured. Commercial arrangements remain dependent on the engagement.
Considerable Potential
In conclusion, Glossop believes considerable potential remains in established operations where processes have evolved without being thoroughly reconsidered. Working closely with customers will always reveal further tasks to automate, but the onus remains on software providers to demonstrate progress and realistic benefits.
For Glossop, the destination is not a logistics operation with zero human involvement. It is one in which people spend less time processing routine information and more time making decisions that matter. The real test of AI will be whether it can deliver that shift consistently in the real world – road closures, snowfall, sick leave and all the other variables that are part and parcel of logistics – with measurable improvements to performance and service.
