Staylittle Telecom is a backup internet and phone system built in partnership with Ryan White for the BBC2 TV Series The Big Life Fix in 2016. Before the system the village of Staylittle was connected to the outside world by only an unreliable phone line and some expensive high-latency satellite internet connections. The Staylittle Telecom system includes a station that connects to the mobile phone (GSM) network, mounted on a disused telegraph pole on a hillside above the village. It is connected via a mesh network to villagers homes, distributed over several square miles.
The system provides a 5 megabit per second internet connection, provided by a single pay-as-you-go SIM card, which is shared by the households that are connected (10 at the time of writing with more planned). All the nodes on the mesh network can be connected to regular telephones which can be used to make calls within the village over the network, or outside the village on regular phone network via a SIP gateway.
When Ryan and I first went to Staylittle I didn’t think we would be able to do very much to help the villagers. Separated from the nearest good internet connection by miles of steep hills and valleys, prone to unforgiving weather, the village seemed impossibly remote. Building and maintaining a piece of infrastructure that even BT with their hundred and two thousand employees couldn’t manage seemed an incredibly daunting task.
With our design research hats on we interviewed the villagers. We found out about their phone system and the satellite internet connections many of them have. I got them to map out who they kept in touch with and how, and to list what they used the internet for and to prioritise those uses.
We found out some surprising things:
The unreliability of the phone system was by far the most important issue. The phone line would go down several times over the course of a typical winter and it could take BT anything up to three weeks to repair it. At this time the village felt totally cut off from the rest of the world. They would miss important news from loved ones and were terrified about needing to call the emergency services and being unable to do so. In one case an older couple had their house flooded at a time when the phones were out and only received help because another villager spotted the flood from the road.
The satellite internet connections provided a acceptable headline connection speed however they had very high latency (the delay between sending an instruction and receiving the response) due to everything coming and going *via space*. This meant that they could not serve as a backup for the phone system (using Skype for example) because the delay made conversations unintelligible. The latency also dramatically slowed down internet research (because of the delay each time a new page was requested) meaning students would need three times as long to do their homework as their peers outside the village.
We left the village knowing that we would have failed if we were not able to provide an acceptable backup for the phone system.
We were dimly aware that there was a ‘bottom-up networking scene’, focused mainly on remote parts of the developing world, so we started researching and experimenting with these mesh networks – where, instead of relying on a central piece infrastructure, everyone is connected directly to their neighbours and information is automatically routed to its destination, potentially using multiple paths.
After some fairly frustrating experiments with Raspberry Pis, Ryan discovered Mesh Potatos – dedicated mesh network nodes, ruggedised to survive challenging environments and with built in support for regular telephones.
The Mesh Potatos arrived speedily from Hong Kong and worked brilliantly straight out of the box. Very well done Village Telco – you guys have built an absolutely brilliant product! I did some testing in my office and then we did more at the Reading Hackspace. After Ryan plugged one into his car battery and we made a call between the hackspace and a random car park over a hundred metres away we were feeling pretty confident about them.
Then we went back to Staylitle.
Walls and mountains
We arrived confidently with our Mesh Potatos and beautiful black and red telephones. We set them up in some nearby houses, tried to make a call and it didn’t work at all. We realised that, unlike the buildings of Reading, the houses of Staylittle are built from solid stuff – thick stone with very poor electromagnetic transmittance.
On top of this the houses themselves are spread out over several square miles of hillsides and valley. Even the ones that have line of sight to one another are often quite far apart and there are some, like Amanda’s, that don’t have line of sight to any other houses.
Luckily Ryan had borrowed a pair of Ubiquity Networks Air Grid dishes. When pointed at one another, over a distance up to 20 kilometres, these create a high-bandwidth link using WiFi radio. By connecting these to two Mesh Potatos we were able to make a call between two houses.
The problem was, who would want one (or more) of those dishes bolted to the side of their home?
We realised that to work out the connections we needed to make we needed to get a proper handle on the landscape and where everyone’s houses actually were. I downloaded some of the OS data for the area and converted it into files that we could cut with a laser cutter. We ordered some smart green polyethelene foam from Kitronik and cut them on the Reading Hackspace laser cutter. With it all stacked up and glued together each 5mm layer of foam translated to 10m of elevation giving us a 1 to 2,000 scale 3D model of the village. I 3D printed some tiny houses to mark where everyone lives.
Having a 3D representation of the village made planning the network much easier. We realised that the longest connection we were going to have to make was less than 2 kilometres which meant we could start looking at the smaller siblings of the Air Grid. Nanostations looked like they might fit the bill so we ordered some for our next visit.
We now had a combination of technology that stood a decent chance of connecting the people of Staylittle to one another, but what about the wider world. Connecting our Staylittle mesh to the wider internet still seemed like a pretty tall order.
Discovering the pole
We had one glimmer of hope. When we had done the original interviews with the villagers several of them had said that when the phones were out they would drive up the hill out of the village and there was a lay by where they could sometimes get one bar of signal on the EE network.
As we didn’t have anywhere near the budget to lay any cables this seemed like by far the best shot we would have at setting up an internet connection.
We got a 3G router that allowed different antennae to be connected to it and put an EE SIM card in it and attached a big patch antenna. I used MIT App Inventor to quickly write an Android app that, when connected to the router would attempt to connect to Google and record the time taken to get a response and the GPS coordinates. I also showed visually when it had a good connection. We taped the router and antenna to the top of the car and began driving slowly through the hills above the village.
We drove all the way out of the village, over the hill and well into the next valley before we got anything like a decent connection which was disappointing. To build a bridge between the mobile network and the Staylittle mesh would require multiple intermediate stations on hillsides and a hill top, all of which would need power – meaning solar panels and wind turbines, caravan batteries, unreliability, huge expense and maintenance, so much maintenance.
We drove back into the village feeling pretty pessimistic. Then something remarkable happened. On the way back in, at a point on the hillside, my app showed a solid connection. We realised that we hadn’t seen it on the way up because the antennae was pointing in the opposite direction (because the car was). We booted up Skype and tried to call Jude.
We had a conversation with him, with decent audio quality, over the internet, at a point on the hillside where we had line of sight to several of the houses in the village. Even better than that we were stopped right next to a disused telegraph pole, and less than a hundred metres from a house (belonging to Nerys) that had mains power. We couldn’t believe our luck.
We would be able to build a station on the pole that could connect to the EE network using a big antenna and to the Staylittle mesh using some Nanostations. We would not need to worry about power as we could run a cable from the house.
This was the first point where I believed we might actually be able to build something that would get the villagers their backup phone system.
Building the network
We knew that to build this network we would need to be drilling holes in the sides of houses, attaching things to their roofs and to the top of a telegraph pole, and messing around with mains electricity. We needed tools, ladders and people who knew how to use them safely. We found John (a satellite installer) and Simon (an electrician) and booked them for the install.
Ryan built up a box to mount on the pole. Inside he put the 3G router, a mesh potato, the power supplies for two Nanostations, a Raspberry Pi (to monitor the system) and power supply for the whole lot, with a 36V input.
He and Simon installed a mains to 36V supply in Nerys’s garage and ran a cable out along the fence. Using 36V ensured that if a sheep chewed through the cable it would not be harmed. John installed two antennae for the router and two nanostations on the top of the pole.
Because Ananda’s house and the Rock Villa are around the side of a hill and do not have line of sight to any of the other buildings, or the pole, one of the Nanostations had to point to Nea and Endaf’s barn – the only structure visible from all three locations – over a mile across the valley from the pole. The other Nanostation was aimed at the roof of the village hall just visible across the hillside.
With John, Simon and Ryan working on the pole I was down in the village trying to work out where we could install other nanostations. In one of the houses I was struggling to connect because of a double-glazed conservatory which was blocking all the signals. I asked the owner if there was anywhere outside where there was power. He replied, “yes, there’s power in the well”. The well turned out to be ideally sited with line of sight to the pole, barn and several other houses. It became one of the key nodes on the Staylittle Mesh.
The TV studio had planned a big party to celebrate the commissioning of the network at 4pm on the Saturday. By the Friday night the pole was complete but we were way behind on the installations and we could not make the connection to the barn and hence any of the houses. We messed around with the configurations until after midnight and went to bed feeling dreadful.
We go up the next morning at first light, ditched the TV crew and drove back to the pole. We listened to Ilmatic to calm ourselves down and Ryan started systemically going through the configurations on each piece of equipment to find the bug. Eventually he found it and after a visit to the barn internet began to flow across the mesh and around the village.
We have been back to Staylittle twice since switching on Staylittle Telecom to do additional instals and hand over the running of the system to the villagers. We have been able to make good quality phone calls out of the village from everyone’s home we have connected (either with phones or on Skype) and the villagers seem impressed by the speed of the internet connection compared to the satellite ones. Ex policewoman Chris is now in charge of administering the system and we are hopeful that it will provide a useful amenity for the village for many years to come.
Aside from the joy of working with someone as incredibly talented as Ryan, and having the privilege of being welcomed into the amazing Staylittle community, the project has impressed some thing else upon me.
Throughout my life I had always associated technological development with driving the centralisation of infrastructure and wider economics – fewer organisations controlling larger and larger networks. It feels like this trend goes back at least as far as the telegraph, maybe further. Working with the Mesh Potatos and Ubiquity Networks hardware – both byproducts of the smartphone supply chain – and building Staylittle Telecom has made me realise that a community can genuinely build its own telecoms network. That smartphones themselves, whilst designed to connect to centralised networks are just as happy connecting to decentralised ones. The entire ecosystem of cheap, low power computers and digital radios (phones, routers and other devices, now numbering many billions) is fundamentally agnostic. If the centralised communication network were ever to fail we might all be able to switch our devices over to a meshing protocol and keep on going. Hopefully it will never come to that, but it is nice feel like that we are heading to a future where we will have more choice and control over the technology we depend on.
The Big Life Fix with Simon Reeve can be seen on BBC iPlayer