A Meter Away - #ameteraway

Inspiration

Inspired by a weekend hike at Macritchie Reservoir (Singapore) with an unusual sight of crowd which leads to unavailability of social distancing.

Crowd Check Inspiration Sudden influx of people in sporting areas to practice a healthy lifestyle or sudden influx of people travelling to supermarkets to stock up groceries can lead to unavailability of social distancing.

Private Check-in Inspiration Since Social distancing is the utmost important now, there was a user experience consideration where people might not want to use the Facebook’s Check-in feature to “check-in” publicly to prevent from being scolded / mocked at to be travelling out of house, hence the private check-in was inspired.

With the above, A meter away - #ameteraway was inspired to create an app for live crowd checking to ensure that locations that individuals are travelling to are not overwhelmingly packed and for the community to have self opt-in private check-in to help with the accuracy of data - emphasizing the importance of Social Distancing.

What it does

Below are some of the supported feature in the chat-bot, note that not all features were showcased in the video,

Crowd Check (Live) Users are able to check the crowd density of the designated location. Upon checking they’ll receive the live crowd situation in the forms below and if the location has places of interest or attractions, it’ll also provide the crowd situation of the top 5 attractions.

E.g. On average Sentosa is - a little crowded and ok to travel to.

Below are the crowd level for top attractions there,

S.E.A. Aquarium -not crowded, safe to travel there
2. Universal Studios Singapore - a little crowded, ok to travel
Siloso Beach - very crowded, not safe to travel
Palawan Beach - quite crowded, not advisable to travel
Fort Siloso - a little crowded, ok to travel

Private Check-in To check in to locations privately at indicated date & timing.

Add location If locations are not available in the DB, users are free to add locations

Leave a feedback For users to leave a feedback for us to notify us of any errors and to better improve the application if they have suggestions

How We built it

On Server Side

We build REST API using Python Flask and Flask-Restplus. The documentation of the API is documented in Swagger.
The REST APIs handle requests such as getting the location lists and registering user's check in information into the database.
The code is then deployed to Heroku and thus the API can be used publicly.
The data is stored in PostgreSQL in Heroku.
The code does not depend on any platform but it is deployed to Heroku as it is free / cost-effective.

On Client Side

As Facebook Messenger is one of the widely used messaging apps, we make use of Chatfuel to build the Chatbot for Messenger.
As most of the users already have Messenger installed, the user would be more willing to use the chatbot to check in / check for the crowd.

Challenges we ran into

Getting people to use - As the solution is "community-based", it depends on how willing people are to share their location information to others to get more accurate data. As everyone is united to overcome the COVID-19, we believe there would be enough people to use it such that it is useful.
Lack of data - One of the initial intuitions was through Facebook’s Check-in feature to use their check-in as well as Google’s Popular time’s data to power our Crowd Check feature however both APIs are not freely available, hence we can only rely on the self Check-in. Nonetheless, we hope for companies with the right technology out there to support this ideation and help realize this with readily available technologies.
Lack of APIs - most geo APIs required a paid plan (e.g. Google’s Places API) hence we’re unable to support globally - also an “add location” feature was added in lieu of this challenge.
Simplifying Idea for Minimal Viable Product (MVP) - Because there are so many possibilities to this idea but our team is only made up of 2 person, one of the challenge was to pick only a few key feature for a MVP that can be completed via the hack deadline
Finding Emoji - It takes us sometimes to find the unicode of the emoji to be returned from the REST API because there are a lot of them.
Learning new tools/apis/tech - To learn new tools to help build the chat-bot

What's next for A Meter Away - #ameteraway

Now that the chat-bot’s beta version is live below are some of the future improvements we have in mind Getting buy-ins of large corporations & making this an info accurate reality :D

Improving the overall application - having more exception handling etc. making the app more robust and fail safe
Introducing more security / encryption features - to better protect the datas collected
Integrating to more chat platforms - expose to more users.
Broadcast important news to the users - for example, if location "A" has people who are infected, everyone should know about it and avoid going there for sometime.
Time sensitive crowd checking - allow uses to check crowd at a certain indicated time
Working with local supermarkets, malls etc. - Since travelling to supermarket is required but we don’t want to have overcrowding situations at those locations, we’re considering maybe even working with local malls, super markets etc. to place sensors on entrances to help power the accuracy of those locations and to advise their customers to check the crowds or indicate their travelling times for better crowd control
Self Opt-in auto check-in - so that users don’t have to keep manually checking in when they are on the move