There are plenty of automated last-mile delivery solutions on the market. But they’re not all equal: some offer far more valuable functionality than others that look similar at a first glance.
How do you choose the right one for your business then?
Well, that’s what we’re going to explore in this blog…
Key Takeaways
- Automated last-mile delivery solutions are not just about route optimisation (or robot cars). They need to address wider operational efficiencies, customer satisfaction, and integration into the wider supply chain
- Efficiency encompasses factors as diverse as dynamic routing, failed delivery workflows, vehicle specs, consideration of traffic conditions, and driver communications
- Customer Satisfaction is another facet to consider when looking at automation: software can help ensure consistently good experiences, by keeping customers informed, reassured, and by being flexible to their needs
- Automated last-mile delivery solutions also need to fit into your wider supply chain arrangements seamlessly
Why Automate Route Optimisation and Delivery Management?
When we talk about automated last-mile delivery, we’re not talking about drones and robot cars.
Indeed, with Amazon Prime anticipating a per-package cost of $63/£48 for drone delivery by 2025, these solutions are best classified under “science fiction” for the time being…
What we’re talking about is technology to solve various aspects of the last-mile delivery problem.
And with last-mile - that is, delivery from the distribution centre to the final customer - constituting 53% of the total cost of shipping, it’s a problem that has a big impact on companies’ bottom lines.
Traditionally, this has been thought of as a variant on the classic Vehicle Routing Problem, which asks:
“What is the optimal set of routes for a fleet of vehicles to traverse in order to deliver to a given set of customers?”
However, there is far more to delivery management than just getting vehicles to destinations as quickly as possible, using as little fuel as possible, and travelling as short distances as possible.
For delivery businesses today, “optimal” includes a lot of other variables that affect operational efficiency - which an automated last-mile delivery solution needs to accommodate alongside route planning.
And efficiency is not the only consideration. Automated solutions need also to consider:
- Customer satisfaction: Ecommerce had been growing fast in the years before Covid-19 struck, but it was the pandemic that transformed expectations around speed, reliability, and quality of service in delivery. Businesses whose last-mile operations provide a poor customer experience will not attract repeat customers
- Integration with other supply chain solutions: ordering, inventory management, and warehouse operations have all been through technological revolutions similar to that which is now sweeping last-mile delivery. Maximising efficiency relies on seamless transitions between all these activities
In the rest of this blog, we’ll unpack those three main areas where automated last-mile delivery solutions need (and please excuse the pun!) to deliver.
Automated Last-Mile Delivery and Efficiency
The Vehicle Routing Problem is, in fact, one of the most difficult types of problem to solve in computational science. Once a certain number of variables are included, the challenge of solving a VRP using “brute force” methods of working out every combination to find the best becomes unmanageable.
So most route optimisation software providers use a variety of heuristics and machine learning, alongside computation. That’s one of the reasons why different software tools come up with different ideal routes.
There are many route optimisation algorithms in commercial use and it’s hard to compare them as they are usually proprietary and always highly complex.
So, our guide to choosing the best automated last-mile delivery tools will focus largely on everything else that affects efficiency, rather than the algorithms themselves.
Routing
As we’ve said, route optimisation isn’t the whole of the efficiency question - but it’s a big part of it. Even the best despatcher will struggle to match route optimisation software for speed and accuracy in designing multi-stop delivery routes.
And on top of that, true efficiency depends on a lot of other factors as well:
- Geocoding: Ever followed your sat nav to an address only to find yourself outside a back gate with the front door around the other side on another street? Well, that’s what poor geocoding can lead to. Different software tools use different geocoders, some of which are better than others
- Sophistication of routing options: Many tools can only handle simple “depot-deliveries-depot” routing. Others, like eLogii, can do far more. For example:
- Long-distance multi-day routes
- “Return to depot” functionality - where a driver visits the depot mid-route
- Custom start and endpoints
- Pickups without corresponding drop-offs
- Multi depot operations management
- Dynamic routing: Basic routing solutions set a fixed sequence of stops that can’t be changed until the route is completed. But dynamic routing allows dispatchers to add or remove stops or change the order while vehicles are in transit. When conditions change, the optimal solution will change too - and the delivery business is one where decisions need to be made in response to changing events fast
- Failed Delivery Workflows: What does a driver do if somebody’s not in when they arrive? According to Loquate, in 2021 6% of UK and 8% of US online orders failed to be delivered on the first attempt - at a cost of £11.60 or $17.20 per failed order. That is a massive cost, which route optimisation alone can do nothing to reduce. An automated last-mile delivery solution needs to have answers to these “reverse logistics” questions as well
- Task templating: Automation doesn’t just help with deliveries in the field. It also saves time back at the office. When choosing a last-mile solution, make sure that it allows you to create templated tasks so that recurring jobs and their special requirements don’t have to be keyed in or uploaded time and time again
Fleet Usage and Vehicle Specifications
If the VRP is complex for a single vehicle, then what is it when you’re trying to optimise for an entire fleet?
And once again, there are other factors to consider:
- Vehicle characteristics: Different vehicles can carry different loads, in terms of weight and volume - and their fuel consumption rates change as their loads increase. Certain kinds of products need to be transported in vehicles with particular capabilities (eg refrigeration, tail-lift for heavy deliveries) while some types of vehicles will be excluded from or unable to perform well on certain routes (eg heavy lorries in crowded city centres). Your automated last-mile delivery solution needs to be able to allocate loads to vehicles in an optimal manner
- Traffic conditions: On paper, the fastest way to the destination is over that bridge. But what about the roadworks on the bridge? And what about the rush hour traffic? Many of today’s route optimisation solutions include traffic conditions as a parameter, and the best will incorporate live data to recalculate optimal routes when hold-ups are reported
- Live vehicle tracking: Maximising operational efficiency often depends on knowing exactly where all your drivers are at any given moment in time. Some older apps (for example, Routific) don’t provide live data - and only show where vehicles were at their last delivery check-in. Without accurate, up-to-the-minute tracking of your fleet, it’s exceptionally difficult to make good decisions about dynamic routing
Driver Management
Automated last-mile delivery solutions also need to help you manage your driver workforce.
Minimising overtime, accounting for breaks, driver preferences, and skills, and territories should all be accounted for in designing optimal routes.
A driver with a damaged shoulder will struggle to unload heavy packages - these should be assigned to operatives who can manage the loads, for example, for the best operational efficiency.
But it’s also important for drivers to be able to feedback information from the field that can help improve performance.
Most tools do this via mobile apps, although some also use dedicated hardware - such as hard-wired GPS, OBD, and/or Electronic Logging Devices - as well.
Data of this kind may include notes on deliveries (eg about neighbours who are willing to take in packages when the customer is out), live communications with control, vehicle telemetry (eg time spent in/out of the vehicle), and more.
Finally, your solution should be able to collect data on driver performance and collate this into dashboards - so you can keep track of who is doing the best job.
Automated Last-Mile Delivery and Customer Satisfaction
Shipping impacts customers satisfaction, according to 98% of shoppers. And Customer satisfaction matters a lot:
So, the more your last-mile solution can standardise and improve the experience customers receive upon delivery, the better it will be for your business.
The days when most businesses could get away with giving customers an eight-hour drop-off window and then leave it up to them to rearrange delivery if they happened to miss it are over!
Customers expect deliveries to be convenient, and they expect to be kept informed of any change of plans - on their own terms.
Your route planning software should include the following as standard:
- ETA calculations for each stop
- Customer notifications around ETAs and delivery windows
- ETA recalculation in response to on-route events
- Notifications of changes to delivery ETAs
- The customer’s choice of communication channel (SMS, email, etc)
Some solutions charge extra for notifications (eg Route4Me) while others only offer a limited choice of channels (eg Onfleet can only provide text messages).
Proving Delivery
Proof of Delivery (POD) is a key part of any logistics operation. Businesses need to be able to verify and prove that their customers have received their orders.
This is another area where drivers’ mobile apps play an essential role in last-mile delivery automation - as these can allow POD data (geo-stamped and time-stamped to provide an audit trail that cannot be tampered with) to be sent directly from the field back to the system.
Most solutions support some basic types of POD collection - photograph, customer signature, etc - but these do not lend themselves to the wider supply chain improvements we’ll look at in the next section.
When choosing a software solution, select one that supports barcode scanning (or QR code scanning).
- Barcodes help drivers to confirm package identity at high speed
- They can make handling returns much less time consuming
- Barcodes allow goods to be tracked throughout the entire shipping process, helping with inventory management, quality assurance, and more
eLogii supports all of these POD collection methods, as well as cash on delivery confirmations.
Delivery on Demand
If your business offers hyperlocal, fast-turnaround, on-demand delivery (eg fast food restaurants) you should be aware that many route optimisation platforms do not support this kind of activity - and some that do are not able to work with third-party logistics (3PL) drivers.
Automated Last-Mile Delivery and The Wider Supply Chain
Delivery doesn’t exist in a vacuum. Last-mile is the final stage of the logistics process, and it should not be considered in isolation.
There are two types of solutions here:
- Select a service that doesn’t just incorporate the last-mile delivery element of logistics - for example, LogiNext and Bringg provide a number of integrated solutions in their software suites. However, these comprehensive supply chain solutions tend to come with a very high price tag, as we’ve discussed in earlier blogs
- Alternatively, opt for a last-mile solution that can be integrated with other tools via an API and webhooks
The vast majority of route optimisation tools provide access to an API, but it’s important to note:
- Not every app provides access on every product tier (eg Onfleet excludes API access from its entry-level service)
- The quality of documentation varies, which can have a significant impact on the need for developer time (eg Bringg’s documentation treats “task”, “delivery”, and “order” as interchangeable terms, and several links are broken at the time of writing)
- Newer APIs tend to be much smoother to work with, which saves you time and money at the point of integration and removes the need for troubleshooting downstream. Here is the documentation for eLogii’s best-in-class API, for example.
The Bottom Line
That’s quite a lot to bear in mind when weighing up the automated last-mile delivery options:
- Can the tool you’re considering maximise operational efficiency gains? Not just in terms of the fastest routes, but across all other factors as well?
- Does it provide functionality for ensuring a consistent customer experience?
- And how easily will it fit into the rest of your supply chain tech stack?
Fortunately, there’s one software service that ticks all of these boxes…