[REPORT]
What route optimization to meet the challenges of the parcel delivery industry?
In recent years, the parcel delivery industry has experienced major changes, driven by the surge in e-commerce sales. As a last mile delivery service, the need for route optimization is often necessary for the large number of packages handled. However, for this industry, the challenge is quite daunting and many parcel delivery companies have been disappointed by previous route optimization software they’ve tried out.
The challenge is so complex that software publishers are struggling to tackle the issue. What are the specific characteristics of the parcel delivery industry? Why don’t route optimization software meet their needs? How can we get around their limitations and solve these highly complex issues? These are the questions we will try to answer in this article to finally provide a solution.
An approach specific to parcel delivery services
Parcel delivery services involve transporting one or several parcels under 3 tons in less than 48 hours. Different types of parcel delivery services exist for every client business needs. This transport’s specificity lies in the platform operations that enable goods to be transported from one end of the network to the other.
The structure of a parcel delivery network includes different stages:
- Pick-up routes: the parcels are collected from senders and sent to the local sorting platform.
- Bundling: goods are unloaded, grouped and sorted based on their destination.
- Haulage: drivers maintain a link between several branches of the network.
- Unbundling: the goods are once again unloaded and sorted into different delivery tours.
- Delivery routes: the drivers deliver the parcels to the recipients and use the opportunity to make pick-up tours.
These operations are at the heart of the process and require an extensive distribution network as well as efficient and reliable cooperation between the different platforms. In such networks, transport operations, especially last-mile deliveries, are often outsourced. This allows for greater flexibility and better control for the parcel delivery— outsourcing the delivery to the recipient is a common practice in the parcel delivery industry.
Route optimization and the challenges of the parcel delivery industry
The nature of the parcel delivery industry makes it incompatible with conventional route optimization techniques
Delivery routes are a major part of the parcel delivery business. Every day, parcel delivery companies deal with huge quantities of packages that they have to distribute among different drivers. The latter can make a hundred or so deliveries every day. In this context, route optimization becomes essential.
However, several problems arise within the scope of parcel delivery companies:
- Just-in-time processes: when deliveries arrive in the morning at the delivery agency, the parcels are unloaded and scanned, then sorted based on their final destination, but these steps are not done one after the other: unloading, scanning and sorting are all carried out simultaneously, and the entire process can take up to two hours. Therefore, it is impossible to wait to have scanned all the parcels to be delivered before starting to sort them. This operation rules out the possibility of using static route optimization softwares, which requires having all the data before launching the calculations.
- The transport partners of the parcel delivery companies are entrusted with a territory that they manage autonomously: the partners are assigned a defined territory in which they are in charge of deliveries and pick-ups. There is no hierarchical relationship between the delivery drivers and the parcel delivery company. Subcontractors are given full autonomy on both the hardware they use (choice of vehicles) and on the process they follow (recruitment of operational staff). Thus, service providers have the freedom to create their own delivery routes, without having any kind of management process forced upon them by the client (except for delivery times). The contracts between the parcel delivery company and their transport partners follow this practice and are generally only called into question if the transport service is not satisfactory or if the service provider is not entirely comfortable with the activity that it has to manage in the area. Keeping a driver in the same area over time can also have advantages in terms of efficiency and quality of service because of the driver’s better knowledge of the area and of the customers’ needs. If it is not possible to modify the routes operated by the drivers every day — overall and daily route optimization is therefore not relevant and can hardly be implemented.
Branch organization through territorial sectorization
Parcel delivery players have found a solution to this headache: territorial sectorization. When parcels arrive at the depot, the dispatchers will sort them based on the delivery addresses assigned to their sector, using the same sectorization system each day.
In the field, each driver is assigned a geographical area, consistent every day, in which they do their tours. Drivers from the same service provider work in neighboring sectors, so if there is an imbalance (low or high activity in a sector on a given day), the drivers can exchange parcels in order to have more equal tours. 24% of the parcels are thus moved from their initial tour to a new tour each morning depending on the day’s activity.
Once the parcels are sorted and the routes are balanced, the drivers load their trucks depending on the route they chose (route planning is often up to drivers).
In reality, sectorization is often a limitation for the implementation of a route optimization solution. Setting up a solution that does not take into account this operational approach would inevitably lead to major organizational impacts: redesigning depot management, reconsidering how transport partners work together, etc…
There is currently no sectorization solution adapted to the last mile for large quantities. Parcel delivery providers are still struggling when they have to change the way their territorial division has been structured over the years. Kardinal took on the challenge of finding a solution to this complex problem without disrupting how depots are organized.
The TAO solution to meet the challenges of parcel delivery
After months of research and development, Kardinal has developed TAO (Territory Analytics & Optimization), its new strategic sectorization solution for the parcel delivery industry.
As a decision-making tool, it provides branch managers with the visibility they need to manage and optimize their territory’s activity.
Based on the use of historical agency data by Machine Learning, “Territory Analytics and Optimization” makes it possible to define the optimal geographical sectorization and to size the related vehicle fleet as accurately as possible.
The tool thus enables parcel delivery agency managers and operations managers to
- audit the performance of a sectorization currently in use
- optimize operating costs by calculating the optimal sectorization for a given territory
- refine a sectorization and assess the impact of local modifications on the sectors
- monitor business activity on a map to identify past hot spots and plan ahead resources
- prepare for the end of year peak period
- simulate changes in the vehicle fleet (new electric vehicles, cargo bikes, etc.) and monitor their impact on costs
An innovative and unique approach to sectorization
This unique approach on the market is specifically designed for depot management and brings in the notion of robustness, an indicator that guarantees the long-term viability of sectors. This concept ensures the creation of sectors that minimize overruns (and therefore the need to rebalance loads between routes before trucks leave the depot). This parameter is taken into account using a probabilistics approach based on the latest advances in applied mathematics.
The “Territory Analytics & Optimization” solution is based on the division of the territory into “cells” representing territorial units calculated from the agency’s historical data (as exhaustive and representative as possible) in order to:
- ensure consistent business activity levels between cells.
- comply with the topography and constraints in terms of road infrastructure
- take into account customer constraints (e.g. zip code).
The sectors are then designed by grouping cells together using optimization algorithms in order to minimize operating costs while providing a minimum level of strength.
Strategic sectorization solution dedicated to the parcel delivery industry
Combine sectorization and route optimization for better management
Kardinal provides a route optimization software, and the “Territory Analytics & Optimization” tool is natively connected to its route optimization tools. Depending on the needs and available resources, it is therefore possible to link sectorization and optimization in a single and integrated approach.
Depending on the desired (and feasible) integration, it is possible to implement different optimization levels:
- Level 1 “dispatch” (this level only refers to sectorization): The Kardinal API processes the incoming parcel flow in real time and assigns each parcel to the right dock based on the sectorization that has been selected and calculated in the “Territory A&O” tool. Drivers create their tours without algorithmic assistance.
- Level 2 “dispatch + sequencing”: the Kardinal API performs the dispatch operation described in step 1 while producing sector-optimized sequences. It works like a round trip where each parcel is assigned to a sector and sent back in a sequenced manner to generate a first version of the tour, which drivers can adjust as needed.
- Level 3 “dispatch + sequencing + rebalancing”: at this level, it is a real-time route optimization, subject to sectors. Sectorization and optimization are fully connected and integrated. The algorithm performs global dispatching and optimization, while integrating any rebalancing between service providers. Thus, each parcel has a priority driver (the one of the sector that covers it), and optional drivers (those of the provider). Thanks to continuous optimization, you can freeze all or part of the parcels in the routes, as you go along, and let the algorithm make its optimized recommendations in real time. At this level, the tool takes care of the entire dispatch and route optimization.
Based on new advanced mathematics and Artificial Intelligence techniques, Kardinal’s TAO solution overcomes the limitations of traditional route optimization solutions. Based on historical data, it analyzes it in depth to provide the best optimization recommendations. Adapted to the specific needs of the parcel delivery industry, Kardinal’s TAO solution supports its operators in their daily tasks and helps them bring even more value to their business.