A(I) Fork in the Road

Welcome to the very first edition of the First-Hand Last-Mile, your monthly dose of authentic perspectives and solutions from experts and peers tackling last-mile delivery challenges firsthand.

In this edition, we look at what lies in store for last-mile delivery and how AI might shape our industry’s route.

Last-mile at the crossroads…

At its core, last-mile delivery is a data industry. From optimizing routes to forecasting demands, tracking utilization rates, and keeping a close eye on the customer experience, business-defining insights are there for the taking. All we need to do is look into the data, make meaning out of it and put it to use (as if). There are so many all-the-time moving pieces, that making meaning out of it feels at times like an impossible task. Not only are you competing against the clock, you’re competing against millions of data points.

So as AI has been taking center stage, the question is will it help make these insights less “static”? Can we hope that in the months and years to come, thanks to AI, routes will be so optimized and demands so well forcast, that there will be no more missed deliveries, no more subpar routes, and no more clogged lockers?

“AI adoption in last-mile delivery has been quite limited, and the industry has been slow to integrate it into actual workflows. While some solutions do use machine learning for data management and optimization, these are not yet aligned with the latest LLM breakthroughs of AI advancements — meaning the real transformation of logistics through AI is still ahead of us”, says Andre Veskimeister.

In his view, the biggest impact of AI on OOH delivery — one of the emerging topics in the last mile — is that it is “set to transform last-mile delivery by shifting customer communication to intelligent conversational agents, reducing reliance on costly call centers while enhancing flexibility. This will give customers more control over their deliveries, enabling real-time parcel redirection and seamless two-way interactions with logistics providers.”

…and customer at the wheel

As we took stock of different perspectives, the customer’s perspective consistently came up as a tenet, where AI has the potential to fully transform the current landscape.

For PostNord’s Christian Østergaard, it is the “entering of personal AI agents that can potentially change the landscape, especially when it comes to brand loyalty. In the not-distant future, we, as end-users, will have multiple AI agents taking care of the time-consuming tasks that we have as individuals — like ordering a parcel.”

“The way I see it, we will be able to give the orders about product specifications or price, but especially the convenience of the delivery, which might be the most important factor. This will impact the carriers wide and far because, all of a sudden, specific delivery instructions will be the most important for the end customer, and thereby, to the shippers as well”, says Christian.

As for the postal perspective and more tangible use cases for AI, in Christian’s view, one of the key challenging issues AI can help with is locker compartment occupancy.

“For us as suppliers, it’s important that all lockers are in use and that they have good rotation. You can compare it to a restaurant — you need to have people on all the tables and if you can have two or three served per table per evening, you’ll be successful.

Today, many parameters determine a good locker usage, like fill rate, pick-up rate per customer, open or closed model, and returns. All these parameters intertwine but also differ from locker to locker based on location, demographics, etc. So, it’s easy to imagine that you need help to find the right solution in that regard.”

Driving the balance sheet

Lowering operational costs is consistently on top of the agenda for the last-mile industry, so it’s to be expected that beside AI’s yet-to-come impact, market players are also mindful of how it can help put operational inefficiencies in order.

For Omni Llama’s Ahmad Abaza, “the most important thing within the OOH is controlling the cost, in order to be able to offer competitive rates for using the network, especially in the new markets. To do so, OOH companies need to operate with light operating modules (Low OPEX) and so, AI will have an impact in automating the work of many departments and reducing the need for employees to handle day-to-day tasks.”

In Ahmad’s view, “our vision into utilizing AI capabilities will be focused on reducing our OPEX within (CX, Locations, CS, Sales, Operations) and to have efficient on-ground operations when it comes to helping manage our lockers capacity, advanced results expectations for the network, scout management, forecast preparations, shippers performance report and production analysis.”

“The efficiency brought about by the AI revolution offers cost optimization for companies that embrace this technology. By predicting demand patterns and automating resource allocation, businesses can achieve substantial reductions in operational costs”, says Marius Gheorghe of Sameday.

And it goes beyond the immediate benefits. As Marius sees it, “perhaps most remarkably, AI is enabling businesses to scale their operations with ease. Complex logistics networks that once required extensive manual oversight can now be managed with sophisticated AI algorithms. This scalability has proven to be particularly valuable during peak seasons and unexpected demand surges.”

‍

This is an excerpt from the 1st edition of the First-Hand Last-Mile newsletter. To continue reading, please subscribe.