Over a relatively short period, a transport or fleet manager’s ability to support an efficient distribution network through route and delivery optimisation has shifted from almost nil to almost limitless.
Indeed, the transition has taken place so swiftly that some companies may still need to fully grasp the present or future possibilities to exploit distribution performance as a competitive advantage.
Whether that is the case or not, a brief study of the origins and progression of fleet route optimisation, covering the last three decades or so and looking forward to the near future, highlights in a spectacular way how technology, in particular, has revolutionised the business of logistics.
So that’s the route we’re taking in this article, exploring the evolution of fleet route optimisation from a time-consuming pen-and-paper exercise to a high-tech process that, in some cases, can be completed in minutes.
Back to the Nineties
To begin our journey, I’d like to share the following anecdotal account of a business acquaintance who spent several years working as a route planner for a distribution operation during the mid to late 1990s.
Simon’s experience as a driver of large goods vehicles engaged in multi-drop beverage deliveries eventually enabled a move into the transport office of the company he worked for.
Simon’s employer was a supplier of beer, wines, spirits, and soft drinks to retail outlets in the United Kingdom. While the primary distribution operation was relatively simple, mainly involving line haul transportation from production plants to various regional distribution centres, the secondary operation comprised multi-drop routes comprising up to 30 deliveries per vehicle per day.
He spent his first few years in route planning at a distribution centre just outside London, responsible for deliveries across Southeast England, including the dense urban conurbations of London and other city and rural customer communities.
Complex Route Planning in the 1990s
Simon undertook the daily planning of several hundred deliveries, all with a lead time of 24 to 48 hours (about 2 days) and relatively tight delivery time windows of three to four hours. Each vehicle in the fleet, typically operated by a crew of two, would either distribute a daily load of between ten and 30 deliveries into areas some distance from the DC or complete two to three local trips, fully loaded, per day into the Greater London area.
As you can imagine, Simon’s task was complex, and at that time, many companies, his included, had no dedicated route optimisation software to help planners complete their work. Here is how he explained the daily planning task to me.
“I would begin my shift at 2 p.m. when telesales would have captured many of the daily orders from customers. We did have some digital assistance in the form of a rudimentary kind of ERP system that enabled us to print ‘planning slips’ which were printed representations of each order, showing the customer details, delivery address and postal code, number of items on the order, the types of packages, and the total weight of the goods, so it wasn’t quite pencil and paper planning. Nevertheless, the process was all manual once those slips came off the printer.”
Pigeonholes and Paper
“We had a system of pigeonholes behind the planning desk, with one pigeonhole for each general postcode area. So, after printing the planning slips, we would place each slip in the appropriate pigeonhole. This process would continue until the order cutoff time at around 4 p.m. when we knew it was safe to start planning without too many new orders coming in to disrupt the task.
The next step was to take the planning slips for each postal code, or two or three postal codes if we knew they were geographically close together, and sort them into piles, representing nominal routes, totting up the weights of the orders to see when a route was maxed out. That’s when the orders reached the maximum weight a vehicle could legally carry.
Next, I would have to fine-tune the planning by shuffling orders between the loads until I had ‘optimised’ the routes for that area—on a purely subjective basis—meaning that I had to be the sole judge of what was optimal.
Then, I would start on another postal code, or group of postal codes, and carry out the same exercise, and so on, until I had assigned all the planning slips to routes. In reality, it would take several rounds of fine-tuning to get all the orders sorted into routes without leaving any vehicles underutilised.”
No Algorithms, But Plenty of Atlases
“The whole route planning process would take several hours, after which I would use the pseudo-ERP system to replicate the piles of planning slips into ‘shipments’ in the system and assign them to vehicles in the fleet. That would enable the printing of manifests and proof-of-delivery documents by shipment, to be issued to the delivery crews the following day, and the picking documents for the warehouse.
This planning process relied heavily on the planner’s geographic knowledge, which I gained slowly through experience, and the only aids I had were A-Z map books of the areas within the DC’s delivery region.
During busy periods such as the build-up to Christmas, it was common for me to work deep into the night trying to fit all the deliveries onto the fleet without forcing drivers into a position where they would run out of driving hours. I also had to ensure that I planned each route in such a way as to make it possible for the delivery crews to meet the customers’ delivery time windows.”
Lots of Planning—Little Optimisation
“To summarise the objectives of the daily planning process, they were, first, to fit all the deliveries on as few vehicles as possible—this was considered preferable to spreading the loads out and using every vehicle on the fleet. Secondly, to ensure all time windows could be met, and, thirdly, to ensure I was not planning to breach drivers’ hours legislation.”
“With no digital technology to assist the route planning process, I had to estimate all route timings, and I could only use weight to gauge when a vehicle would be filled. However, that was not a very reliable indication because often, when it came to the actual loading, the volume of the vehicle load space would max out before the weight.
I had no way to consider the nature of the roads and streets that the vehicles would use, so there could be no route optimisation by fuel consumption, kilometres driven, minutes elapsed, or any other objective parameter. I guess that’s why we called it route planning and not route optimisation.”
The Arrival of Digital Tools
“Fast forward to around 1998, and my company procured a software application which effectively turned the physical pigeonholes into digital ones and removed the need to print and sort planning slips, but it did little more than that, and it was still a case of shuffling around digital versions of the planning slips. The application had no on-screen mapping or anything like that to help visualise the plans.
After another few years, we upgraded to a new route-planning system based on digital maps. This application included automated load and route planning processes. However, the system’s algorithms were very rudimentary, and therefore, the routes it created required substantial manual tweaking to make them realistically achievable.”
So that was Simon’s experience in route planning, 1990s style, which no doubt reflects how many companies went about the process back then. Even today, some companies are planning distribution fleet routes using legacy applications that are not a patch on the solutions now available.
If you believe yours is one of those organisations, now might be the time to review your route optimisation maturity to see if a technology upgrade would be beneficial, or simply to explore best practices and identify improvement opportunities.
If so, the team at Logistics Bureau is always ready to assist you. We don’t sell software or represent any companies that do. Instead, we offer impartial consultation, advice, and practical help to determine and execute steps to improve your fleet route optimisation capabilities, and if necessary, to select tools that best match your needs.
Ready to level-up your route optimisation? Book a free consultation.
We’ll fast forward again now, and look at how route optimisation works today, some 20 years on from Simon’s time as a route planner.
Fleet route optimisation has moved on a great deal since the turn of the 21st century. As I recall, over these 20 years, the software available to fleet managers graduated from digital pigeonhole shuffling to basic map-based routing applications.
Next came more sophisticated map-based systems utilising detailed parameters such as average speeds for various road types, mandatory drivers’ rest periods, and a standard-minutes approach to calculating work time.
The next generation of solutions could also plan and configure loads based on packaging dimensions and the capacity of vehicles. Finally, we arrived at what we can consider today’s industry-leading software, which we’ll look at now in more detail.
Converging Technologies Change the Game
Fleet route optimisation has recently been significantly reshaped thanks to the convergence of several key elements. The latest fleet management platforms enable transport and fleet managers to optimise fleet routing for cost, productivity, customer service, fuel efficiency, and other commercially desirable factors.
So, what converging technologies are we talking about that make today’s fleet route optimisation a much faster and more effective process than anything offered in the first decade and a half of this century? Let’s take a brief look at some of them.
GPS and Telematics
Global Positioning System (GPS) technology and telematics systems provide real-time data on vehicle location, speed, and driving behaviour, allowing transport managers to track fleet vehicles and gather data that they can use to manage their drivers’ performance.
It also allows them to communicate proactively with customers in the event of vehicle maintenance issues, heavy traffic, or other conditions that might delay delivery.
Integrating these technologies with route optimisation software takes fleet management up another notch, because you then get a single view of what’s happening, in reality, to compare with what you planned to happen.
If you are a route planner, you can use the data generated by GPS and telematics systems to inform your planning decisions, gradually making route plans more realistic and likely to be followed by your drivers.
Big Data and Analytics
With the latest analytics technology, you can further refine your fleet’s efficiency and plan more accurately. Today’s analytics platforms and the accessibility of data generated from various sources mean you can exploit information originating not only from your company’s internal systems, such as telematics, if you have it, but also from external sources. That enables you to account for a broader spectrum of variables in your planning activity.
AI and Machine Learning
When you bring artificial intelligence into the route optimisation equation, things get especially interesting. Indeed, route planning applications with integrated AI are arguably the pinnacle of excellence in current route optimisation technology.
Machine learning algorithms are the final link in what I like to call closed-loop route optimisation, integrating planning applications with vehicle telematics, GPS navigation, and analytics software. Such a system can enable planners to create routes and transmit them directly to drivers via an in-cab display or mobile device.
Drivers’ adherence to routes is visible in real-time, and planners can even change a driver’s itinerary on the fly, in response to a long traffic tailback, for example. Meanwhile, all real-time data from the fleet can be fed back and processed by machine learning algorithms to inform the route optimisation engine, allowing the system to adapt to changing conditions, predict traffic patterns, and optimise routes dynamically and autonomously.
The Ultimate in Fleet Route Optimisation
When all the technologies described above come together, it’s possible to plan and optimise routes comprising hundreds or even thousands of deliveries with far less human input than those 1990s solutions we explored previously.
Furthermore, the planning process is reduced from hours to minutes, enabling order cutoff times to be delayed or even eliminated—especially since incoming orders can be added into routes as they are captured—and the entire plan tweaked and adjusted accordingly, as many times as are necessary throughout the planning period.
In such a scenario, a primary constraint could be the ability of the warehouse operation to keep up with dynamic route planning. But with routes being adjustable while vehicles are on the road, drivers can even be rerouted into distribution centres, local warehouses, or retail outlets to pick up additional orders (in operations where this is feasible).
Would you like to cut down the time and effort involved in fleet routing? Our experts will be pleased to help you identify, procure, and implement an appropriate solution to save you time and money, and deliver the optimisation performance your enterprise deserves.
Do you have route optimisation challenges to overcome? Talk to our experts.
The Future of Fleet Route Optimisation
It’s hard to see how technology can improve much on the current ability that software gives us to create efficient route plans.
We already have automated load and route optimisation systems that can interact with onboard software in vehicles, guide drivers along planned routes, and receive and analyse data returned from the fleet. With the addition of machine learning, these systems can use that captured data to continually improve the accuracy of planning to reflect reality more closely.
So, what more can be done from a planning-software perspective? The degree of planning automation can still increase, further lessening the need for human intervention. But how can fleet routing be further optimised once the entire circular process is fully automated?
The answer could be to focus on the execution side of route optimisation. Indeed, we are already seeing advances in the distribution arena that will eventually change the entire scope of possibilities.
To conclude this article, let’s briefly explore a few potential developments that might perceivably enhance—or even wholly transform—fleet route optimisation in the next few years.
Smart Infrastructure
If there is one final summit in the rise of smart distribution route planning, it could be the point at which a company’s route optimisation software integrates with municipal traffic management systems.
As governmental bodies increasingly advocate for and invest in smarter cities, even the infrastructure of transportation routes is gaining digital communication and data-sharing capabilities.
When corporate route-planning and execution platforms begin to integrate and communicate with these municipal applications, we may see another step-change in distribution fleet asset control and management, supporting smoother traffic flows, reduced congestion, and an overall improvement in fleet efficiency.
Autonomous Fleets
Across all the fleet route optimisation elements that I’ve touched on in this article, there is one common factor—the presence of humans in the operation of distribution and delivery assets.
To date, all but the tiniest handful of vehicles (all experimental) engaged in distribution require a driver. However, the future could see significant developments in commercial vehicle autonomy.
Autonomous fleet assets will likely be the biggest disruptor ever seen in the route planning and optimisation arena, and they will probably take several forms. Let’s look first at the possibility of autonomous surface vehicles.
Autonomous Trucks and Vans
If the day finally comes when trucks and vans can operate safely and effectively without human drivers, there will no longer be a need to consider rest and break periods to limit the workload of these assets.
If linked to the intelligent infrastructure mentioned above and simultaneously to route planning systems, these vehicles will endlessly ply our highways and streets. They may even be capable of loading and unloading goods autonomously, or at most, requiring only the intervention of loading personnel at the point of origin and human resources/MHE to unload at the point of delivery.
The actual delivery method, of course, would depend upon the type of distribution concerned and the willingness of customers, for example, to execute physical unloading if necessary.
In the case of home deliveries, for instance, autonomous vehicles might be equipped with smart systems to eject parcels for retrieval by the customer. Meanwhile, in the case of commercial freight, customers might unload their goods much as they do in today’s warehouse receiving operations.
At this stage, we can only speculate about the details since, for now, they are not necessarily foreseeable. Still, history has shown us time and again how technology has turned the seemingly impossible into practical reality.
Autonomous Alternatives to Trucks and Vans
On the other hand, we may see a divergence in the types of assets used in distribution fleets. For parcel delivery, the vans, motorcycles, and bicycles we see today might be replaced by small robotic vehicles capable of negotiating streets and sidewalks to carry packages to homes and similar delivery locations.
At the same time, commercial distribution could remain the realm of large goods vehicles, albeit operating on an autonomous or semi-autonomous basis.
In any such scenario, route optimisation would be far superior to what’s currently achievable. However, if there is one step further that we can go, it is towards a concept that is already a reality, on a limited basis, in various parts of our world—the shifting of distribution from the surface to the skies.
Road-free Route Optimisation
Drone delivery operations have the potential to remove all man-made and natural obstacles from the distribution landscape, eliminating the need for anything other than straightforward, point-to-point routings from distribution sites to map-based coordinates, which drones will probably be able to figure out for themselves on the fly.
That said, the need for safe air corridors may mean that drones will be forced to make routes that are not quite “as the crow flies.” Nevertheless, like higher altitude air traffic today, which is bound to follow airways and accept a certain degree of routing from air traffic control, drone route management is likely to be considerably less complex than in terrestrial transportation.
Could Drones Ease Road Congestion?
Of course, it’s unlikely that distribution-by-drone will become practical for shipments large enough to require freight handling (although entrepreneurial projects exploring the use of large UAVs for just such a purpose do exist). More realistic is the assumption that drone use for door-to-door parcel delivery will become a mainstream transportation mode.
That will be all to the good if control systems achieve sufficient sophistication to support a high volume of low-altitude drone activity, since not only will routes for the drones be straightforward and efficient, but by taking some distribution away from road networks, it might also enable more streamlined routing for conventional or automated surface transportation.
Again, everything I have written here is pure but reasonably informed speculation. Still, it should not seem so farfetched when you consider that the idea of drone deliveries could easily have been dismissed as pie in the sky only a few short years ago.
Yet here we are, witnessing a changing logistics landscape in which delivery drones are beginning to see acceptance—and some efforts at enablement and regulation—among transportation governance bodies worldwide.
Is Your Fleet Routing the Best It Can Be?
In this article, I have summarised some of the highlights of the rapid evolution of fleet route optimisation over the last 30 years or so. Whichever way you look at it, the technological advances have been remarkable, and look set to continue apace. To wonder what route optimisation will look like in another 30 years is enough to boggle the mind.
Of course, every company with a distribution fleet to manage will have achieved some degree of maturity in the route optimisation journey.
It’s hard to imagine that many still need to shuffle paper documents around using pigeonholes. However, some perhaps use unsophisticated—but adequate—digital tools for route planning. In some cases, active optimisation may be unnecessary, especially for companies that primarily operate long-distance trunking or have predictable routes and schedules.
But where there is a requirement for dynamic routing, it pays to invest in the best optimisation technology your company can afford, as the savings delivered by effective route planning are substantial.
People Still Matter in Fleet Route Optimisation
Even with the most advanced solutions available for route optimisation, there is always a need for some degree of human intervention, and that becomes more critical with less sophisticated systems, so don’t underestimate the value of experienced route planning staff, who know, for example, how to adapt to routing vehicles within new geographies.
Finally, do you believe your company should improve its fleet route optimisation? Are you setting up a new distribution operation, or do you have plans to upgrade to a new route-planning application? If so, advice, support, and hands-on help from an external company like Logistics Bureau can be immensely beneficial.
Our people have been helping companies to optimise their fleet routes for decades now, and while any route optimisation program will probably be a one-off or infrequent event, we do it all the time. With our assistance, you can complete your projects faster, with fewer frustrations and less expenditure, and achieve the best possible results.
Need impartial advice and support with fleet routing? Book your free consultation now.
Editor’s Note: The content of this post was originally published on Logistics Bureau’s website dated November 22, 2023, under the title “Fleet Route Optimisation in the Past, Present, and Future“.
