Route optimization is the process of finding the most cost-effective route for a set of stops. Many people think this means finding the shortest distance or fastest time between point A and point B, but this is not entirely true. Route optimization is used when you want to minimize drive time for multiple stops, while also accounting for a range of complexities like customer time windows, vehicle capacities, driver schedules, and more. In this column, I do my best to get you involved in this critical discussion which has been concentrated deeply among experts all over the world.
Route planning helps field service companies and delivery businesses plan out the best routes for their drivers each day, whether they are trying to provide reliable ETAs and improve customer satisfaction or get through a multi-stop delivery route in the most efficient way possible. Well-planned routes mean your drivers spend less time driving, which reduces fuel costs and can increase both time onsite and the number of stops a driver can make in a day (1).
Maybe the first question is that how route optimization works. Route optimization typically uses algorithms. The reason for this is because the complexity of route optimization means that humans can’t easily compute all the different parameters to find an optimal route, especially in a short amount of time. Route optimization algorithms aim to solve two of the most difficult computer science problems: The Traveling Salesman Problem (TSP) and the Vehicle Routing Problem (VRP). For instance, imagine that you’re a salesman and you have to visit a bunch of cities. Let’s say 100 — because you are a very ambitious salesman.
Now how would you go about doing that?
What is the shortest possible route you can take between those 100 cities before returning home?
It’s actually a really difficult problem to solve — almost impossibly hard. Computer scientists call it a NP-hard problem. The Traveling Salesman Problem gets exponentially more difficult to solve the more dots, or cities, the salesman has to visit. Now if you thought the TSP was difficult, consider this: What if it wasn’t just one salesman you needed to plan routes for? What if you had multiple salesmen — a whole team of them? Now you need to decide how to split up the visits and optimally assign them to your team. That’s an even more challenging conundrum known as the vehicle routing problem or VRP. Mathematicians and computer scientists have been trying to solve these problems for years. The closest we’ve come, albeit imperfectly, are algorithms. Up until recently, only big corporations had the budget and resources to implement this type of technology. But now even small businesses can access complex algorithms via simple, easy-to-use delivery route planners or route optimization apps. Navigation apps like Waze, Google Maps, and MapQuest might be the tools of choice when it comes to planning a simple route and getting turn-by-turn directions from point A to point B. But what happens when you need to plan routes for point A to Z? What happens when you want to plan for an unlimited number of stops or have multiple stops to plan – 10, 100, or 1,000 destinations? The optimization of stop order is one way in which delivery route optimization software separates itself as a category from Google Maps’ software, which functions as a route planner for multiple stops, but not as a route optimizer. (1)
I think now is a good opportunity to get you more familiar with some software and platforms that are utilized for routing optimization.
Verizon Connects route-planning software is one of the software that have been utilized to help dispatchers and route planners calculate the most cost-efficient routes for their drivers quickly and efficiently. It also gives them the tools to respond to changes and update already planned routes in near real time as situations change throughout the day. Integrated into their fleet management software, their routing software also allows fleet managers and sales reps to improve driver management and plan for future business changes (2).
Another platform is Route4Me that has over 32,000 customers on almost every continent. Route4Me's mobile Android and iPhone apps have been downloaded over 2 million times since 2009. Extremely easy-to-use, the apps synchronize routes, enable two-way communication with drivers, offer turn-by-turn directions, delivery confirmation, and more. Behind the scenes, Route4Me's operational optimization platform combines high-performance algorithms with data science, machine learning, and big data to plan, optimize, and analyze routes of almost any size in real-time (2).
Another essential question that should be answered is about why route optimization is vital for the AVs operations.
The first important matter relates to the limitation in power range. Long distance travel is one of the biggest challenges of autonomous vehicles for deliveries. Although it is still unclear how much electricity self-driving vehicles will use, most experts agree that just the onboard equipment could consume up to 2,500 watts per second, depending on the vehicle’s configuration (3).
The second issue relates to the inefficiency of AVs to distribute a large number of units. Autonomous vehicles for deliveries typically cannot fit a lot of units and will only be able to deliver three to five units at one go. One way to overcome this challenge is by adopting routing optimization software to optimize the routes for the driverless delivery vehicles. When you have multiple units, each unit takes up capacity with each item having different dimensions. Routing software can factor in all these constraints, while planning and distributing the routes. So, it can ensure that all your self-driving vehicles are optimally balanced, for example fitting in two large units in one car, while assigning five small units to another. Advanced route planning software like Route4Me also come with multiple parameters, such as the maximum pieces per route to set a limit to the number of pieces a vehicle can carry as well as the maximum weight and volume per route to set the weight and volume constraints for each vehicle (3).
The third matter attributes to the issue of traveling through traffic. Using autonomous vehicles for deliveries makes sense if you are delivering in the countryside or cities with low-density areas. But, what if you need to deliver during peak hours in Manhattan or Chicago? Can you imagine what a mess that could be? So, you need to ensure that your autonomous vehicles do not travel through high-density areas and do not get stuck due to bad weather, left turns or U-turns. Doing this manually may be impossible. This is why it is best to use a route planner. Route planning software can help you optimize routes for thousands of stops in less than a minute with almost 100% accuracy. It also factors in traffic, weather, and other constraints. In this way, your autonomous vehicles will not get stuck, deliveries will be made on time, and your customers will be happy (3).
In sum, routing optimization plays a critical role in today’s transportation. Moreover, the future advancement in mobility will highly depend on the time and budget that are allocated today in order to develop routing optimization in smart manufactures and intelligent infrastructures. We should not forget that without paying enough attention to the subject of routing optimization, reaching the short-term and long-term targets in smart mobility will be impossible.
(3) blog.route4me.com/a-closer-look-at-the-challenges-of-optimizing-routes-for autonomous-vehicles-for-deliveries/