The Things Network sets up a world-wide Internet of Things (The Things Network)

The Things Network (TTN) is an initiative started by Dutch civil society. It aims to have LoRaWAN networks installed in every city around the world. By interconnecting these local networks, TTN wants to build a world-wide infrastructure to facilitate a public Internet of Things (IoT).

Smart everything

LoRaWAN is one of many Low-Power Wide-Area Network (LPWAN) protocols that can be deployed to build an IoT network. What all of these protocols have in common is that they support (very) low-bandwidth communications over longer distances: for example, sending just a handful of bytes every four hours over five kilometres. These technical characteristics allow the use of (cheap) remote devices that can last five to ten years on a single battery.

Note that LoRaWAN networks are very different from other wireless network infrastructures like those used for WiFi and mobile phones. LoRaWAN is just another wireless technology, featuring its own specific characteristics, and catering to its own specific applications.

Typical IoT applications involve networks of remote sensors, actuators, embedded systems, and moving objects like (self-driving) cars, bikes and others assets. Hence the name "Internet of Things" and references to smart cities, smart power grids and smart roads. Industry experts expect the number of devices that will be connected to the IoT in the coming years to be in the billions.

Policy Context

You are the network

Last year, we ran into the LoRaWAN technology at an Amsterdam hackerspace meeting, says Wienke Giezeman, the initiator of TTN. With a reach of up to ten kilometres, we could cover the entire city and cater to thousands of devices using only a dozen access points ["gateways" in LoRaWAN parlance] costing 1000 Euro each. That could really be a breakthrough for a lot of IoT applications.

After seeing how telecom providers are currently building IoT networks — closed, based on proprietary protocols, and using a subscription-based revenue model — Giezeman and his business partner Johan Stokking decided to take a different approach: We wanted to build this network in a crowdsourced way, getting citizens, businesses, universities and developers to build the networks and connect it all together using open source software. You know, like how the Internet was built.

The ambition was to build an open, distributed and decentralised IoT network, crowdsourced, owned and operated by its users. After pitching this idea at an IoT Meetup in Amsterdam, more makers became interested and a manifesto was written.

Take off

Within six weeks, TTN had managed to find locations for the access points in Amsterdam and have them connected to its server. The gateways are placed in co-working spaces, media businesses, startups, the Port of Amsterdam, and two "big four" accounting firms. Some of the businesses involved are contributing additional services to help the initiative succeed.

At the same time, a team of programmers worked on a server system to process the data sent and received through the network, thereby forming the glue between the LoRaWAN network and the IoT applications.

Within a week after the launch, TTN had received several dozen requests from other cities willing to join the network. One year later, the network comprises 200 connected gateways spread all over the world.

Description of target users and groups

Giezeman aims to facilitate other people and cities world-wide to set up their own access points and networks. He does not, however, want to get involved in actual implementation projects. Every helping hand we offer at one place takes away initiative from someplace else. So we create a context for people and organisations to install a LoRaWAN network themselves.

Description of the way to implement the initiative

Crowdfunding

Since the costs of the LoRaWAN gateways are often covered by individuals who want to set up access points, the relatively high price of these antenna systems has proven to be a barrier to further growth. That is why in October last year a crowdfunding project was launched on Kickstarter. It aims to develop a LoRaWAN gateway costing only 250 Euro.

One month later, almost 300,000 Euro had been raised by over 900 backers, making possible the development of a cheap access point (the Things Gateway) and two types of development kits (the Things Node and Uno, the latter being based on Arduino).

According to Giezeman, the hardware will be ready for shipping this summer. We are currently working on the production and distribution of the gateways, and sourcing all the required components. This part of TTN has been placed in a separate entity named The Things Products.

When we are ready, we will ship four thousand devices to one thousand people. That will increase the size of our network tenfold in a single step.

Technology solution

Technology choice: Mainly (or only) open standards, Open source software

Main results, benefits and impacts

According to Giezeman, the next step will be to create value on top of the new network infrastructure. Users can make existing business processes more efficient, or create completely new applications. The power of the Internet is that it facilitates extreme diversity, bringing together individuals as well governments and businesses. Everything you could ever wish for is or can be made available on the Internet.

Business cases

Giezeman uses the "smart mousetrap" as his basic business case example for the use of the IoT. If you put down mousetraps, you have to check them every day to see how many mice you have caught. That means you have to make a round of all your traps. But if you could enhance these traps with smart technology that notified you when each trap was sprung, you could go directly to a specific trap, and only when necessary.

A solution like that would require only a little data traffic, and only on a few occasions. Such a trap would be just 10-20 Euro more expensive than an ordinary mousetrap.

One step further, of course, would be to use the data generated by the traps to find out where the mice are hanging out. But we tend to 'move forward looking backward'. The first television shows, for example, were basically radio programmes with images. The real revolution will come only after the technology is in place.

Peak shaving

Of course there are more realistic IoT business cases available, Giezeman continues. For example, you can use sensor networks for "peak shaving" on the power grid. Using alternative energy sources like the sun and wind generates more extreme highs and lows in available power than we used to have. These peaks and troughs will have to be eliminated using traditional power generation technologies. More information on the consumption of energy can help you optimise supply and demand.

The Amsterdam sewer system, for example, contains sensors that signal when the pipes and drains are full and the pumps should be activated. What if we could postpone this action until late at night, when we can buy green power from Germany at negative prices? That is another case that requires an extensive sensor network to collect the data needed to make these kinds of decisions. Ger Baron, CTO at the City of Amsterdam, is very open to such ideas.

We also use our own tools to build applications ourselves, as show cases, and because as a developer you have to 'eat your own dog food'. We want to make the development of IoT applications as easy as possible. That is why we are making available not only an affordable access point but also these Arduino-based development kits.

The TTN website presents several actual use cases implemented on the Amsterdam TTN network.

Return on investment

Zero bills

The costs of the infrastructure are currently covered through our initial funding, says Giezeman, but we do not rule out the idea that participants and users will eventually have to pay for our services. If you take more value than you give, we will provide 'options to compensate that'. A typical example would be if you send or receive a lot of messages through the network. We are looking for a model where everyone contributes.

That way we want to avoid a 'tragedy of the commons' [where a shared resource is depleted by users acting in their own interests rather than that of all]. So we are transparent about our costs, and we will start sending out zero bills just to show participants that maintaining a network does cost money. In addition, we are currently talking to several potential partners about funding.

Track record of sharing

Open source

All software and hardware related to TTN is or will be available online as open source. According to Giezeman, the server software is not yet 100 percent decentralised, but he assures us that it will be soon, as stated in the development roadmap.

To assure the independence of this new distributed network infrastructure, TTN is currently talking to ICANN and RIPE — who are responsible for the technical basis and the name/number spaces of the traditional Internet — about finding a good home for the central (root) node of the network.

Lessons learnt

For governments, TTN offers an opportunity to bring maker communities and businesses together, says Giezeman. According to him, the cost of the LoRaWAN gateways is not the largest problem when setting up a new local IoT network. It's finding the locations to install these antenna systems. That is where governments can play an important role: they generally have a lot of locations spread all over the city.

Scope: International, Local (city or municipality), National, European, Regional (sub-national)

Categorisation

Type of document
General case study

Attachment

Documentation
Login or create an account to comment.