Universally Challenged - Space Apps Challenge

Universally Challenged| Human Factors

Human Factors

The emergence and spread of infectious diseases, like COVID-19, are on the rise. Can you identify patterns between population density and COVID-19 cases and identify factors that could help predict hotspots of disease spread?

AiPart

Summary

AiPart uses real world data and A.I. to simulate how people move through any given space, which can be tweaked to obey social distancing rules and more, and then generate a floor plan or path to allow people to return to as close to normal lives as possible, while remaining safe.

How We Addressed This Challenge

Our Goal: How can we optimise any generic space (city, hospital, school, shop, airport etc.) such that the most popular routes taken by visitors, maximise social distancing, minimise social contact and minimise the time spent in that space. This approach could supersede “Contact Tracing”, given that prevention is better than cure.

People don’t like “being tracked” so we offer a solution that “pushes people through the system, rather than forcing a system on people”, where optimisation is applied to environmental spaces, rather than trying to micromanage many individuals’ movement directly.

Apple designed their HQ in California to increase the chance of serendipitous meetings.
We propose the exact opposite, but with a similarly positive goal.

Our project looks at population density globally to find areas with high infection rates, and then we can target these areas, and the spaces within them where infection is most likely to occur. One of the many factors affecting the spread of the virus is how people move. We’ve decided to address this by having artificial intelligence decide where is best for people to be able to walk in any given area, while keeping them away from other people, and minimising their time spent in what could be a risk area.

AiPart can essentially be applied on any scale, the whole way up to a full city! This would act somewhat like Google Maps for pedestrians, while also keeping them away from other people, rather than just simple directions from here to there.

How We Developed This Project

We chose this project because we felt this challenge’s solutions could have the greatest worldwide - and beyond - impact. Our approach to creating and designing our project started with some brainstorming! We wondered how we could get people back into the world but still keep them apart from one another and safe?
We looked through NASA Datasets related to population density, and realised that in an area with lots of people, there were lots of Covid-19 cases, by cross-referencing with other maps.
In those areas, people still need to be able to go to the shop and buy essentials, or go to a hospital for treatment, etc. so we thought there must be a way to let this start happening again, without increasing, and possibly decreasing the rate of infection.

Visually when creating our submission slides we wanted to not only focus on what the solution is but also how to present it. We used the ‘Jet Propulsion Lab PowerPoint Presentation Guidelines’ to help show our solution in a professional and effective manner within NASA’s own set guidelines. Expanding on this visual side of our solution we wanted to ensure we were constantly tying every element together so we made sure to focus additionally on the smaller details like logo colours and brand naming. This can be seen by our use of blue background to connote the theme of technology, reliability and competence backed up by ‘99 Designs Technology of Color Guidelines’. As well as this, the play on words on ‘AiPart’ sounding like the english word ‘Apart’ relating to our theme of social distancing and space while including the use of Artificial Intelligence in the “(Ai)part” section.

One of us had been experimenting with solving a maze using code, and wondered if we could apply that to any given space. As it turns out, all spaces have the same set of contents:
walls, some entrances, exits and spots within them people want/need to visit.
Now we knew what our goal was, it was time to start developing the solution.

Using Python, Visual Studio Code, Github, and a powerful PC, we were able to simulate the movement of randomly generated people with randomly generated agendas over and over through a space. The next step will be to plug Tensorflow or another A.I. engine into this simulation to tweak parameters and find the optimal solution for any given situation.
The training of the A.I. will be very time consuming, as the more people we introduce concurrently to a simulation, the more computationally explosive it gets.

One of the biggest problems we encountered was trying to condense all the information and ideas we had surrounding our project into just a few slides. We wanted to make sure that we got an adequate amount of detailed information into the PowerPoint, without making it a pain to read. We feel our biggest achievement is the fact that within 48 hours, we created more than just a solution to a problem, but a full plan of action for a program that can be applied almost anywhere.