## Inspiration

We focus on the simple mantra “Now is the time”! Sustainability is becoming more important nowadays. What struck us was knowing about the percentage of CO2 shared by the construction sites to the global CO2. It is huge!

Moreover as Benjamin Franklin says “Lost time is never found again.” This is why mindful planning and work on productivity are so important.

From this fact, we thought about a solution for the challenge “Combatting chaos caused by construction” specifically for reducing the time wasted during construction to help both people who are working in the field and the others who get affected by it.

Not only that, if we could provide a solution to reduce the chaos of construction sites and also reduce the wastage of construction materials, optimize the time of the construction logistics, we can also contribute to the CO2 emission reduction. In short, we are working towards the greater good!

## What it does

Our solution will help the stakeholder visualize the construction site logistics and help them make better decisions. It shows the bottlenecks in their supply chain and gives them feedback on how they could improve it.

Reduced the chaos and cost by

• minimizing the number of Days

• reducing the number of Trucks

In summary, our solution is:

Good enough: this is an optimization problem that is extremely hard in general, however, we approximate it in a way to find a close to the optimal solution which reduces the cost and less complexity in general.

Modular: In a way that it is extremely easy to add and delete the constraints from the encoding.

## How we built it

Mathematical Modeling:

We defined the requirements mathematically as constraints and optimized (minimized) our objectives mainly the number of days by keeping in mind that we should be using reasonably many trucks.

Here are examples of the constraints that we have: Hard constraints (can not be violated at any cost): At a time, a truck can be used only for one purpose Soft Constraints (Incurs cost when violated but it’s okay to violate these times): Don’t use more than n number of trucks (if possible)

Approximations:

It can be approximated to simplify the task of the solver: These problems are inherently difficult because the number of feasible solutions can be tiny with respect to millions of possibilities that may or may not occur. A naive way in which we have approximated is that we considered all the payload jobs will take the same time which is equal to the maximum possible time taken by any job. This can be improved by refining the constraints depending on the need.

We build a Single-Page-Application (SPA) with React using different kinds of components to visualize the transportation of the payloads. React API, etc as front-end and python.

## Challenges we ran into

Front-end perspective: A difficult task was to find out how we should present the result and specifically the effect of our optimizer. It was also a bit difficult to coordinate those tasks because we had remote- and in-person members in the team. Backend perspective: Identifying constraints to obtain the optimal solution is hard and challenging.

## Accomplishments that we're proud of

We could creatively show this optimized schedule helps the stakeholder understand how much time was and is wasted on specific tasks, especially in the transit part.

We could solve a really hard optimization problem that is sound and mathematically proven to be correct within less than two days.

## What we learned

We learned to incorporate new problems with people from different countries. It was interesting to see how other people think about a problem and how they proceed to solve a problem. We didn't have much time so we had to make decisions quickly.

## What's next for Combatting chaos

The Frontend should be extended to make it easier for the stakeholder to recognize the bottlenecks. The data produced from the backend needs to be made available over an API (REST or something like that). The backend is modular to achieve any kind of extensions, approximations, and optimizations.

1. Traffic and Weather Data integration
2. Want to combine the fancy ML? We have room to extend the toll by adding future predicted traffic and weather during the transit time of trucks to and from the concrete plants and Jobsite.
3. Recommender System or Chat Assistant
4. Intelligent assistants to help the stakeholder to understand and analyze the data, logistics schedule and make better decisions using recommender systems.