Transforming the last mile

Marcel Bruyns, Axis Communications

---

I’d say I have a pretty good relationship with my Takealot delivery guy. We know each other by name and I let him park his car in my driveway for a few extra minutes so that he can safely rearrange his boot, making sure that the packages he’s set to deliver next are close to the front and easy to access. I mention this because Henz – that’s his name – has created his own system to streamline the process of actually putting an order in the hand of a customer.

Much like Henz, the broader transport and logistics industry is constantly looking for smart ways to increase efficiency during the last mile of delivery. This stage is the most expensive and time-consuming part of the shipping process; what makes it even more tricky is the fact that it is probably the most important step in influencing a customer’s satisfaction with the experience.

According to Greg Cline, head of corporate accounts at Investec for Business, it’s best to start conversations about efficiencies and cost-savings in the transport and logistics sector with a discussion about the sector’s challenges. Only then can one consider what technologies are needed to solve these problems. Solutions such as blockchain, drones and robotics are important, but the emphasis should be on machine learning and AI, he says, because having access to real-time data is the key to improving the industry’s ability to react to different scenarios.

With AI-enabled predictive analytics, organisations can do more effective forecasting and scenario-planning to meet changing production and retail demands and better plan loads.

Luigi D’Amico, iOCO

Luigi D’Amico, a senior sales executive at iOCO, and Julian Hanmer, a senior consultant for advisory and design at iOCO, agree. They believe that AI can deliver the innovation transport and logistics businesses need, particularly when it comes to planning. “We’ve practically exhausted traditional route planning,” says Hanmer. “But when we use AI, it’s far easier to change the plan if an accident closing a freeway occurs, for example.”

It’s the kind of stuff Sixty60 is doing to improve driver efficiency and reduce delivery times. The Sixty60 app uses route optimisation algorithms to overlay geospatial and traffic data with available delivery time slots so that they can determine the best possible delivery route. They’re also combining customer order data with machine learning tools to outline optimal delivery zones for each store. Again, this is so that orders are allocated and delivered as efficiently as possible.

It’s not just about efficiency on the road. Warehouses are the heart of logistics operations, says Hanmer. “An issue we see a lot among manufacturing and logistics companies is a lack of effective demand forecasting. This is traditionally done using Excel spreadsheets, but if forecasts are flawed, businesses have to bear the costs of holding excess raw materials and they can’t meet spikes in retail demand, or plan loads efficiently.”

“Even the most experienced people can only do so much and process only so much information,” says D’Amico. “With AI-enabled predictive analytics, organisations can do more effective forecasting and scenario-planning to meet changing production and retail demands and better plan loads.”

If you’ve ever tracked a package that was “out for delivery” in real-time, you’ll understand that one of the greatest challenges around the last mile problem is inefficiency. That’s because the final leg of shipment typically involves multiple stops, unsuccessful delivery attempts, as well as delays due to traffic congestion and ineffective route planning. All of this adds to delivery costs.

Many of these challenges are most felt in rural areas, where delivery points can be many kilometres apart. In South Africa specifically, finding solutions to better service markets that are currently underserviced, such as townships or hig-risk or hard-to-reach areas, is important, says MD of Bob Group, Andy Higgins.

Driver buy-in for the use and onboarding of these different technologies is essential, as is understanding driver performance.

Kimberley Taylor, Loop

In these locations, something like smart parcel lockers and counters can make the last mile more efficient, while also providing flexibility to recipients who don’t want to wait around all day for a delivery to arrive. “Interactive delivery management enables recipients to select delivery slots and makes it possible for them to inform the courier that they are not available for delivery. They are even able to change the delivery location if need be,” says Higgins.

While the last mile presents a wealth of challenges, it also presents a world of opportunities for improvement, say D’Amico and Hanmer. A common issue adding to last mile costs is single purpose vehicles making long trips and returning empty. “For example, a container truck delivering a shipping container from City Deep to Rustenburg will go back empty. The same applies to a specialised vehicle such as a refrigerated milk truck – you can’t do a return load because the vehicle must come back empty and be cleaned.” To address this, the industry could incentivise clients who purchase half a truckload of goods to rather purchase a 74 brainstorm.

full truckload of goods at a slightly discounted rate. This will reduce the transport costs associated with empty loads, D’Amico says. “The sector can reduce costs and improve efficiencies by connecting logistics and partner systems to prenotify the destination so that they have loading bays available when the truck arrives. This eliminates congestion and waiting times. The broader ecosystem needs an ML- and AI-enabled logistics system upstream and downstream to support efficiency.”

Something as simple as camera technology can streamline various last mile processes, says Marcel Bruyns, sales manager for Africa at Axis Communications. By incorporating something like in-vehicle cameras, transport and logistics brands can provide proof of delivery. He says Axis currently uses these in-vehicle solutions to record footage from cashin- transit and courier vehicles across the globe.

The industry is also looking to its vehicles to improve efficiencies over the last mile. This, coupled with the growing importance of sustainability, is driving the adoption of EVs, says Higgins. Take, for example, the technology found in electric vehicles or driverless cars, which can be used to influence a driver’s or operator’s behaviour around the best times to stop, accelerate, or what route to take if there is a traffic jam ahead, says Cline. If we were to imagine a fleet of trucks that could upload data in real-time – taking into account the environment and situations encountered on the road – we could create repetitive patterns that inform the best course of action for a particular scenario and can minimise, and even prevent, many of the challenges that cause a business to lose time and money.

In a new study, MIT researchers showed that using machine learning to control a fleet of autonomous vehicles as they approach and travel through traffic intersections improved traffic flows, and reduced fuel consumption and emissions. By controlling how each autonomous vehicle accelerates and decelerates, the researchers were able to eliminate stop-andgo traffic entirely. Unsurprisingly, this had a positive impact on traffic congestion.

While the potential for these technologies is huge, Kimberley Taylor, CEO, founder and the brains behind the Loop delivery management platform, says that buy-in from the driver is critical to managing and mitigating the industry’s headaches. This is a major hurdle, she says. Driver buy-in for the use and onboarding of these different technologies is essential, as is understanding driver performance. You can bring in the most advanced technologies, but if performance is not understood well, you can’t improve it, she says.

And so I come back to Henz. Technology might be transforming how the industry operates, but if a human being is the last touch point on the customer’s shopping journey, their job is crucial. If they are friendly and efficient, they play an important role in making sure that the customer’s experience is a positive one. And, hopefully, that they shop again.

While autonomous vehicles (AVs) are nothing new – early experimentations with driverless cars were happening in the 1920s – a confluence of trends is pushing the evolution and deployment of AV technologies, according to Forrester Research. In the world’s major cities, governments are looking to AVs to reduce traffic congestion and air pollution. Modern passengers are increasingly favouring AVs because they deliver eco-friendly and convenient mobility experiences. Car manufacturers view the advancement of AVs as a positive because they offer an opportunity for them to innovate, stand out from their competitors and give their customers new and unique experiences.

In commercial logistics, the AV push is about resilience, sustainability and improved supply chain management, says Forrester. For example, something like the use of autonomous forklifts or autonomous drones in a warehouse can easily complete several tasks without any need for human intervention. In addition, AVs also hold potential to improve long haul journeys and the frequent stops that are part and parcel of last mile routes. In the future, there is potential for AV technologies to integrate with broader traffic management systems so that routes can be adjusted based on traffic flows and delivery arrival times can be adjusted based on real time information.