How to pick the right use case for your grassroots IoT project

In Germanys Rhine-Neckar area (which is around the cities of Heidelberg, Mannheim and Ludwigshafen) we are currently preparing the setup of a regional IoT infrastructure based on LoRaWan and TheThingsNetwork.

Why LoRaWan?

LoRaWan offers a good network if you are looking for a low bandwidth, low power & high range network solution (2km in a city, and 6km+ outside of cities). This extended reach allows you to connect sensors and devices, which are outside of the network coverage of Bluetooth and WiFi and they are cheaper than putting SIM cards in every sensor.

While there are a bunch of other platform options (3G,4G, Bluetooth, Wifi, Sigfox, NB IoT), LoRaWan allows for the cheap and grassroots based setup of a network of gateways (these devices connect the devices with the sensors), simply because everybody can either purchase a readymade gateway or everybody can build one himself using public tutorials.

LoRaWan especially helps, when you don’t want for your local telco to build IoT network coverage or when you want to set low entry thresholds for users (e.g. when a sensor costs just 20EUR, a battery and you are done)

Why TheThingsNetwork?

TheThingsNetwork gives you a proper IoT platform, which allows you to focus on your data & how you want to consume them. This is done e.g. by enabling users to translate compressed byte values in machine readable JSON structures, which even humans can understand.

Our motivation

We are interested in offering this IoT platform, but we realised, that it might be difficult to gather (even monetary) support, if we cannot tell for which use case we want to build it initially. (However, once the platform is there, everybody can use it for everything, but you need to start with a tangible application for the technology.

Our criteria for selecting a use case

In our brainstorming session, we found the following criteria, which you might find useful for your project selection.

Feasibility (as a project)

First of all you want to estimate whether your use case can actually be done. This goes into two directions:

Your project must be technically possible under the known constraints. While doing a complex computer vision task on an Arduino board with low power is difficult to implement, measuring noise & environmental data is actually doable and you can proceed. The assessment, whether something can be measured, should be done with some creativity and workarounds in mind: e.g. when counting people in an area might be difficult, you might find it easier to count mobil phones (since many people have mobile phones with them now).

The second dimension of feasibility addresses the project framework, in which you are often working. The project character (often as a time, resource limitation) is clear, additionally you often work with volunteers, who need a higher motivation level than paid people.

Being able to also use resources e.g. of the near university, where people do research in your area, might ease these limitations but requires more coordination on the project side.

Utility & public awareness

You often want to do something, which actually serves as many people as possible. In your IoT you could e.g. support local beekeepers as a side activity — since unfortunately there are not that many beekeepers — , but the impact and the public awareness you can generate is limited then.

By choosing a use case, which affects all parts or at least large parts of the population, you will get better opportunities to showcase your work and the public interest will be higher.

Visibility in the public space

A good impact of your project can be reached once people are able to see & ideally are able to interact with your work results.

This can happen on a local URL, which obviously requires to be a able to use the URL and you need to do quite a marketing effort to make your project known.

An concrete idea for visibility in public space might be to show your measurements in realtime in a hotspot in your city on a large display. This might even have some art character, which enables you to also include the local creative scene into your activities.

An additional channel are sensors, which also get data back from the network. Think of e.g. a environment sensor, which show a yellow or red light, once certain thresholds are exceeded.

Area coverage

IoT networks gives you the option to place quite many sensors in your area of interest. Using these options you want to aim of a representative coverage your area and you want to measure at small scale.

If you are only looking to get information about a specific spot in your city or if you need excessive processing power, an IoT network might not be the right platform for you.

Digital & innovative

In case you feel that your city & area are lagging in terms of applying digital technology you might focus your use case selection on something, which only just now get possible e.g. due to IoT technology.

In our case we felt a disconnect between the technical options and the experience level, which large parts of the population have with technology in general. If people additionally distrust technology or what one can with all the big amounts of data, then you might want to focus on a easily digestible application of sensor measurements, which gives more transparency or — even better — gives actionable insight to the public.

Should work in increments

Especially your first use case should very fast deliver some data & some insights. A use case, which only begins to work once you have thousands of sensors & infrastructure deployed, might show difficulties in keeping your team motivated. Additionally you need to make sure, that your existing monetary means can bring you to a level, where you can show results.

Loosely following the Minimum Viable Product method from the Lean Startup framework helps to keep focussed on the first success. And gathering more funds get easier, once you can show a positive track record of things achieved.

Potential to participation

If you want to do a project with public visibility and large impact, you sooner or later will want to enlist support from the public. You are looking then for people, who provide limited time & resources out of motivation to support your project.

Your use case should enable this kind of participation. With IoT projects, you want to check your use case, that you can e.g. provide an open sourced tutorial, how everybody can build a sensor or can work with the opendata datasets generated. A good example is luftdata.info, which includes an easy tutorial, how everybody can build a high quality fine dust sensor.

—

These were the criteria, which we used to determine our two first groups of ideas.

We’ll investigate two ideas further:

• One is gathering realtime data related our living environment, and
• the other one is a public “dash button”, which allows people to interact with the items in their surroundings.