The problem Narad solves

COVID-19 has caused massive destruction as cases pass the 1 million mark. It is a disease which can only be prevented, for the near future. That being said, it is evident that awareness and advocacy are the most important tools we have to inform people about the disease, toll-free lines, symptoms, food distribution centres, etc.

In most first-world countries (the US, UK, etc), this is done through the internet, radio broadcasts or the television. However, each country’s strategy against the coronavirus has to be altered based on its economical, geographical and social diversities. That being said, countries with less geographical, social, and linguistic uniformity (such as India) have varied access to information, especially in rural areas in non-developed countries where such facilities may be sparsely available, usually outdated and inaccessible to a large population who speak vernacular languages.

For more context, India has 122 major languages and 1599 others according to the 2001 census. Additionally, there is inconsistent data about the virus in various places making it very easy for the public to be directed towards misinformation and leading to ineffective deployment/donations in multiple areas.

• A large population
• Wide socio-economic disparities
• Unequal access to resources
• Unequal access to information/Lack of awareness
• Linguistic diversities even among close areas (Nagaland, a state in India, has as many as 15 recognized different languages and 20 different dialects spoken in it)
• Extremely isolated as well as densely populated regions

The current methods of information broadcasting aren’t sufficient or viable for many affected areas especially the isolated rural areas which speak their own dialect/language most of which tend to have a lower literacy rate and reduced access to infrastructure and resources.

Challenges we ran into

We've always been fascinated by microcomputers like the Raspberry Pi, and we were amazed to discover we could underclock the CPU clock speeds of one core to effectively be able to transmit at FM frequencies. With the first version of what we built, we decided to take it a step further and add a simple piece of solid gauge wire, which acted as the antenna. With this setup, we could transmit at distances over 500m with no distinguishable loss in quality.

However, we soon realised a single Raspberry Pi couldn't cover the area of an entire Indian village. As such, we decided to use a mesh (a type of network topology, akin to star networks) system consisting of multiple raspberry pis, with only one requiring an internet connection. We settled on a WiFi based mesh network, since Raspberry Pi Zero Ws are inexpensive and come with WiFi antennas built in. We also tried using Zigbee and LoRa (Long Range radio) for this, but soon realised the extra cost didn't carry sufficient benefits for us, since each unit being relatively cheap is a major selling point for Narad.

Initally we had the plan to have a few Narad Units built with extra features such as live broadcasting in the local area, for village Sarpanchs (the head of Indian villages) to be able to transmit mission critical information whenever required. However, this plan didn't work out as the units would've need a screen(preferably touch based) and a microphone, which significantly add to the cost. This led us to build the Narad app, which gives people in local governments and intuitive way to live broadcast locally relevant information.

The Narad app is built using react native, and communicates with the Raspberry Pis in the mesh network by joining the WiFi network they create.