RouteSync

Inspiration

With the development of technology and industry, problems related to the ecosystem become more and more severe. Inspired by the workshop from Holcim MAQER, we aim to optimize the route of trucks to minimize empty runs. It not only saves businesses precious resources but also nurtures a greener tomorrow.

What it does

We designed a formula to calculate the efficiency of transporting goods based on time and cost per mile driven by a truck. The truck can have partial loads but should prioritize maximizing its capacity without going too far off the planned route. Generally, the outbound trip will be fully loaded, but the inbound trip might have one full or one to two partial loads.

How we built it

To solve this problem, we need to devise an algorithm that can find the most efficient loads for the inbound trip. Here's a step-by-step breakdown of the algorithm:

Step 1: Define Parameters

• Truck Capacity: Maximum volume the truck can carry.
• Load Details: Details of potential loads (volume and location).
• Cost per Mile: Cost incurred per mile driven.
• Time Matrix: Matrix representing the time taken between different locations.
• Max Deviation Distance: Maximum distance the truck can deviate from its path to pick up a load.

Step 2: Algorithm Development

Step 2.1: Initialization

• Initialize the outbound trip with a full load.
• Create an empty list to store the selected inbound loads.
• Calculate the direct route time and distance for the return trip without any load.

Step 2.2: Inbound Load Selection

1. Identify potential loads that are along the return path or within the maximum deviation distance from the path.
2. Create combinations of one or two loads and evaluate them based on:
   • Total volume (should not exceed truck capacity).
   • Total deviation distance (should not exceed the max deviation distance).
   • Total time and cost for the return trip with the selected loads.

Step 2.3: Optimal Load Selection

1. Select the load combination that maximizes the total volume without exceeding the truck's capacity and minimizes the total time and cost.

Challenges we ran into

The mapping, and connection between each part. We used at least three different frameworks and it was quite a big challenge for us to transfer data to each other.

The main challenge was the shortage of time, we came up with too many ideas but could only implement a few of them, which was quite a pity.

Accomplishments that we're proud of

For the front end, we created various customer, cement supplier, and driver interfaces to aid with order tracking, delivery schedules, and route navigation. \ Additionally, the management of the cement factory can adjust the specifications for each type of raw material, which affects how the recommendation system functions and lists the most ideal roundtrip.

To determine the best method for selecting the route, we combed through a large body of scholarly literature. Finally, after putting all of our data together, we are able to determine a favorable round-trip while keeping the raw material inventory percentage. We also left a ton of interfaces for future development, including Deep Learning and other data mining.

What we learned

The most important lessons learned are working with varied abilities, and handling ambiguity. This project simulates difficulties encountered in industry life, developing participants' problem-solving skills and fostering teamwork.

What's next for RouteSync

We can improve the appearance of UI design with professional designers, and build a better user experience.

Also, as mentioned above, the core part of our work is a route recommendation system, which could be easily upgraded with the technology with AI. For this purpose, we left interfaces for future development, even now we have the parameters for seasons, suppliers, etc. We can easily feed more information to the model, and improve the accuracy, and also the degree of automation.

Built With

• expression.js
• javascript
• mongodb
• node.js
• pandas
• python
• typescript
• vue.js