3. Connect a constellation of drones to be deployed during a crisis

Disclaimer: this is a personal side project still in progress

How might we build an interconnected constellation of drones to be deployed amidst a crisis to provide a network and guidance to the population in a hurry while gathering data for authorities?

Client: Personal Side Project - Brazil - 2020
Role: Designer on training :)

CONTEXT

This is a personal project, which I started to build new skills. My inspiration came from the increasing number of crises caused by population growth, climate change, and industrial disasters. Examples are the Brazilian dam collapse of 2019 and the decimating fires in the Amazon, Australia, and California in the past couple of years. During these events, authorities and the population had to act fast to save others and the ecosystem. However, the lack of real-time information and loss of connection are two critical factors that delay people's response time.

In this exercise, I tried to reduce those two factors' impact by creating a constellation of drones to be parked in high-risk areas, like industrial facilities, or transported by the authorities when a crisis arises, like the fires in California. These drones serve two purposes: provide internet service to share information with the local population and gather data for authorities to build better action plans.

This personal project focuses on building hard skills for building technologies that solve real challenges.

CHALLENGE

This project has three key challenges:

Develop an application for people to receive information.
Prototype drones to be used in different crisis contexts.
Create Artificial Intelligence (AI) for interpreting data, and sharing information with the authorities and people.

KEY QUESTIONS

How will the population connect to the application, and what scenarios should be considered for making it accessible?
What are the different types of information the population will receive in the application?
What type of data must the drones gather?
How to make the drones relevant and accessible in different types of crises?
How to make the drones connect in a constellation?
What kind of data must the AI program gather, interpret and share?
What technologies and libraries should be used for making the AI relevant?

GOALS

This work-in-progress is an explorative personal challenge to learn the fundamentals of technologies such as AR&VR, VUI, 3D modeling, physical product prototyping, digital manufacturing, sensors, and AI/ML. The goal is to introduce myself to subjects freely before undertaking a master's, rather than focusing on the final solution due to the complexity involved in achieving it.

Emergency app
Build an emergency app for alerting the population, guiding them to safety, and allowing them to share their status with the authorities.

Drones Constellation
Deploy a drone constellation to gather real-time data regarding the situation and extend an Internet signal for the locality population.

AI for Data Analysis
Develop an AI program for interpreting the drones' data and support the development of an action plan to tackle the issue at hand.

Learning goals diagram

Explorations

Emergency App

The app would be available in regions with high risk for industrial and natural disasters, like forest fires. It solves three problems:

Alert
It sends alerts to people's phones concerning the situation and which actions they must take. Learning goal: design considering human behavior in a crisis.

Directions
It provides the best paths to take, considering blockages and data coming from the drones. Learning goal: to build an AR experience.

Report
It allows people to send the authorities their real-time location, condition, and personal information. Learning goal: build a Voice UI.

So far
I learned about user behavior in crisis and designed the UI. I am learning how to use Unity and Mapbox for building an AR Navigation experience.

Mockup 1. The population receives an evacuation alert. Mockup 2. After clicking on "Go to safety," the AR Navigation shows the direction. Mockup 3. After arriving at a safe place or stopping moving, the Voice UI interacts to learn their condition and share it with the authorities.

Drones Constellation

It would be deployed from fixed towers at high-risk regions or from mobile stations transported to the scene. When a crisis starts, it will:

Mapping
Be equipped with sensors to obtain data. Learning goal: work with sensors.

Constellation
They will cover a larger area as a constellation providing real-time data and serve as internet range-extend antennas for the population to use the app. Learning goal: learn about IoT and connected devices.

So far
I have learned how to build a drone prototype and use a Laser Cutting Machine. Now, I am exploring how to 3D model more complex structures. That is exciting, as it is my first experience building hardware.

This is my first rapid-prototyped drone. To illustrate, the material used: 2.4G Remote Control, Two pairs of Hubsan engines, four propellers, a Syma receiver circuit board, and a Hubsan 3.7v battery.

AI for Data Analysis

(Next step to work on)

This project is an opportunity is to have a complex challenge to learn about Machine Learning to analyze real-time data.

For Authorities
It will provide reports about the situation, which direction the event is moving, where people are located, and forecast the situation.

For People
It will provide the best directions to follow or what necessary action to take.