ConstrAPPI

Two payload status: Delivered, Recalled. The ratio is a measure for success in each city.
On-site traffic in Boston: Payload crowded activities in several days are detected
7 additive types in Boston payloads: The distribution vs Date
Histograms: for 3 cities Batch, Journey percentage, On-site activity and delivery time is plotted.

Inspiration

We were excited to see that construction gathers so much data at various process steps and amazed by the issues construction causes to communities. These issues have been prevalent for a long time now. We were interested in identifying the underlying issues by analyzing the data helping to solve challenges faced by various stakeholders and contribute our solutions.

What it does

We divide the solution into two main parts:

We create an application that construction companies and suppliers can use to get the latest weather updates to know if they should order and supply the concrete to the job site, else the concrete can not be poured. We also provide the construction companies information about the businesses that will be affected due to the construction site causing noise, air pollution, and road closures so that they can be notified in advance by a single click. For the suppliers, we also show which plant should be used to send the concrete/materials to the construction site to use the shortest route.
Most people depend on Google/Apple maps in their day-to-day lives to find the best route to travel around. These maps reactively deal with congestion caused by construction via crowdsourcing, leaving many of their users in traffic jam. We have created a proactive approach to stop this congestion before it begins. We created an API that mapping companies can use to identify congestion causing construction, they can use this data to update their maps and in turn routing algorithms can choose the most efficient route that avoids the construction site. This reduction in congestion will reduce chaos and improve sustainability of areas around constructions sites. Additionally, the bus/taxi companies can also query the API for their own apps like Uber, etc.

How we built it

Data Analysis

  We used Python in Google Colab and R for analyzing the data and visualizing the plots. This was the most important step to understand every underlying issue that happens in this environment. We were able to identify quite a few things by understanding and visualizing the data like one of the variables provided - delivered_at, wastes 5mins to manually mark the delivery is done which can be automated at the truck wash state.

Mobile Application

We use the ionic framework for creating a mobile application. We also used various APIs like google-places, google-maps, and an alternative to google-directions API. For the weather forecast, we use OpenWeatherMap API. We also design in the app a button that needs to be improved to send messages to identified businesses about the construction site and the times of the construction work being done.

Congestion API

We built an HTTP API in GOlang. It contains an SQLlite database with data about timings and locations of cement deliveries. Clients can then query specific cities or building sites to gain the relevant data. We have built the API in mind of growth to allow further developments such as connecting with construction companies to have a dynamic database of up-to-date congestion causing events. As well as moving beyond just cement deliveries to all forms of construction that cause congestion.

Challenges we ran into

The whole team is participating remotely so communication and managing team members in various time zones is the starting challenge. On the technical side, the first challenge was to understand the data and clean it in order to start analyzing it. The second challenge was to find various APIs we can use in the mobile application for getting weather updates, businesses near a construction site, finding the best route from the plant site to the construction site. We also ran into the issue where we had to find an open-source alternate to google-direction API as it is not free for even a single-use.

Accomplishments that we're proud of

We were able to manage time and come up with concrete solutions we think are going to help the stakeholders make the process efficient and sustainable. We made a working prototype and provide an API that can be used by Google or Apple to mark construction confession on their maps. We had a great team who worked together and everyone showed great initiatives.

What we learned

We learned the various stakeholders associated when construction is done. We learned how we can analyze data and provide insights into underlying issues by visualizing data. We polished our skills for mobile app development, building APIs, and data analysis. Working in a team and collaborating without decreasing productivity is an important lesson as well.

What's next for ConstraAPPI

Due to lack of time and data available, we were not able to identify the number of trucks that are available at a plant site to design an optimal model. We also were not able to get the information about the truckload that is allowed on various roads in different countries to optimize the travel process and manage the carbon footprint. We want to gather these available data and design the most efficient model for this process and decrease the carbon footprint as much as we can.