Road Transportation Network Visualization with Python
Build visualizations of your FTL network performance, including deliveries/routes, cost per ton, and truck size.
Road transportation is critical to supply chain operations, representing a significant company cost.
With diesel prices increasing and the ongoing pressure to reduce CO2 emissions, there is a growing need for transportation optimization.
As a data scientist, how can you leverage descriptive analytics with python to support this optimisation effort?
Logistics systems (WMS, TMS) generate a large amount of transactional data, which requires advanced tools to process and exploit.
In this article, we will explore how to build visualizations of road transportation network performance using Python.
- Process and analyze transportation records
- Improve visibility into current routing and truck loading rates
- Simulate multiple routing scenarios to estimate the impact on the average cost per ton
Summary
I. Build a Transport Plan with Python
1. Problem Statement
2. Objective: Reduce the Cost per Ton
II. Understand the Situation with Python Visualization
1. Import Datasets
2. Listing of stores delivered by each route
3. Add cities covered by each route
III. Automate the visualization of Transport Plans
1. Transportation Plan Visualisation
2. Geographical Visualization of Store Deliveries
IV. Next Steps
1. Generative AI - Transportation Control Tower Powered by GPT
2. Measure the Environmental Impact
3. Routing Optimization: Graph Theory
4. Do you want to generate animated graphs?
Build a Transport Plan with Python
Building on our series on Warehousing Operations Optimization, we now turn our focus to enhancing Road Transportation efficiency.
How can we use Python to manage transportation networks using visualization?
We will apply a similar approach by:
- Processing Data: Extracting and structuring transportation records to create an optimization model.
- Improving Visibility: Leveraging Python visualization libraries to clarify routing and truck loading rates.
- Simulating Scenarios: Developing a model to test various routing scenarios and assess their impact on average cost per ton.”
What do we want to achieve? Minimize the cost per ton.
Problem Statement
Retail Stores Distribution with Full Truck Load (FTL)
- 1 Warehouse delivering stores by using three types of Trucks
(3.5T, 5T, 8T) - 49 Stores delivered
- 12 Months of Historical Data with 10,000 Deliveries
- 7 days a week of Operations
- 23 Cities
- 84 Trucks in your fleet
How can we achieve this? Shipment consolidation.
Maximizing shipment consolidation is crucial for reducing costs in route transportation.
We are using 3rd party carriers that charge full trucks per destination:
By optimizing the number of stores covered per route, we can fully utilize larger trucks to lower the cost per ton and enhance overall efficiency.
The table above shows rates applied by carriers for each city delivered for each type of truck.
Observing that costs per ton are lower for larger trucks, we want to cover the maximum number of stores per route.
🏫 Discover 70+ case studies using Python and VBA to automate reporting 🤖 and optimize transportation networks 🏪 in this Cheat Sheet
Understand the Situation with Python Visualization
Import Datasets
Before considering the optimization model, your priority is understanding the current situation.
What data is available for us?
Starting with unstructured data from several sources, we must build a set of data frames to model our network.
Records of Deliveries per Store
Store Address
Transportation Costs
Listing of stores delivered by each route
Let us process the initial data frame to list all stores delivered for each route.
1 Route = 1 Truck ID + 1 Date
Add cities covered by each route.
Let us now calculate Transportation Costs invoiced by carriers
Visualization: % Deliveries per Truck Size
Insights
- Average Truck Size: a large majority of small trucks
- Cost per ton: the inverse proportion of cost per ton and average truck size
Find more inspiration for smart visualizations in this tutorial,
Automate the visualization of Transport Plans
Transportation Plan Visualisation
We want a simple visualisation of all deliveries per day with a focus on the number of different routes.
Solution: Python’s Matplotlib grid function
- Columns: 1 Column = 1 Store
- Rows: 1 Row = 1 Day
- Colour = White: 0 Delivery
- Colours: 1 Color = 1 Route (1 Truck)
Visual Insights
- Delivery Frequency: n deliveries per week
- Number of Routes: How many different colours do you have daily?
- Colour = White: no delivery
- Routes: do we have the same stores grouped from one day to another?
12 Months Overview
After optimization, this chart will help us to visualize the impact of new routing easily.
A better routing means fewer daily routes, so you’ll have fewer colours per line.
Geographical Visualization of Store Deliveries
Let us generate a visualisation of geographical locations delivered in the same route.
Visualization of the different routes covered per day
🏫 Discover 70+ case studies using data analytics to monitor supply chains ✅ and costs reduction 💸 in this Cheat Sheet
Next Steps
Transportation management teams can use these visualizations to pilot their operations and challenge the status quo.
Can we develop prescriptive solutions to support the transportation operational teams?
As we have extracted and formatted data from multiple systems, we can exploit this source of insights using advanced analytics solutions.
Transportation Control Tower Powered by GPT
In November 2022, OpenAI released the first version of ChatGPT.
Generative AI has become an opportunity to improve the user experience of any analytics product using large language models.
I shared my explorative journey of LLMs used to boost Supply Chain Analytics in a simple case study.
The idea was to use LangChain to design a smart agent connected to a TMS database to answer operational questions.
For more details,
What about the CO2 emissions of our transportation network?
Measure the Environmental Impact
In addition to cost reduction, you can also target CO2 Emission reductions by optimizing your transportation network.
Your organization invests resources to build capabilities for sustainability reporting and footprint reduction.
![A diagram depicting the calculation of CO2 emissions for a logistics route. The route involves multiple segments: a road shipment covering 120 km, an air shipment covering 10,000 km, and a final road shipment of 450 km. The total shipment weight is 1,100 kg, and the formula for calculating total emissions is displayed at the bottom: “E(total) = 1,100 × [(120 + 450) × E(road) + 10,000 × E(air)].” This highlights the process for calculating emissions across different transport modes using Python.](https://miro.medium.com/v2/resize:fit:1400/format:webp/0*xPPpDDL_UbBJ7G-I.png)
How can we measure our contribution to the sustainability roadmap?
I have implemented a simple methodology to report the CO2 emissions of your Distribution Network using Python and PowerBI.
More details in this article,
How can we support the implementation of improvement plans?
Collaboration with Graph Visualization
The aim is to improve the transportation plan to increase the average size of trucks by delivering more stores per route.
How can we support the change management?
Because of operational limitations, you are limited by road restrictions
- Delivery Time Window: Stores can receive products only at a certain time of the day
- Road Restrictions: Large trucks are forbidden on some roads
- Unloading Conditions: Some stores need to be delivered first
Therefore, you cannot perform this optimization exercise alone.
A graph is a structure that contains nodes (stores), and each related pair of nodes is called an edge.
As data scientists, how can we improve communication with operations? Visualization.
It is vital to collaborate with the transportation teams that have experience managing route planning daily.
For more details,
Do you want to generate the animated graphs shared in the previous section?
Generate Animated Graphs using Pillow.
An animated map showing the different delivery routes can be generated to help transportation teams compare their solutions.
You can animate your map using Python’s library Pillow to create results like this one.
The idea is to generate each frame and generate a GIF using Pillow.
For more details, have a look at this article.
About Me
Let’s connect on Linkedin and Twitter, I am a Supply Chain Engineer who uses data analytics to improve logistics operations and reduce costs.
For consulting or advice on analytics and sustainable supply chain transformation, feel free to contact me via Logigreen Consulting.
If you are interested in Data Analytics and Supply Chain, look at my website.
📘 Your complete guide for Supply Chain Analytics: Analytics Cheat Sheet
💌 New articles straight in your inbox for free: Newsletter
