Enmeshed in la la land — part 1

Mountain dreams

Audio podcast for this blogpost

Against a grey sky pierced by columns of light, chased by worsening snow, flanked by a himalayan wolf, as our SUV tracked up the Pensi-la, the humbling, never-ending expanse of the Drang-Drung glacier finally revealed itself. A month ago, we had crossed this very spot, in the dead of the night with little clue as to what lay before, or beyond…

****

The long road to Zanskar began, as most do in India, over a chai almost two years ago. I was discussing my experiences of setting up off-grid wireless mesh networks in Ladakh and Spiti with Prasenjit Dey, a close friend and an expert hacker. He spoke about his dreams to visit Zanskar one day, and of the people he knew in the valley, and his idea of connecting remote villages with each other, and to a shared knowledge base like the one that may be provided by the Internet-in-a-Box(IIAB). He discussed this with Geshe Lobsang Tseten who belongs to the Karsha gompa, a 1000 year old monastery in Zanskar, and he was interested in the educational and cultural aspects of the idea. I was instantly hooked, and while the trip didn’t materialize in 2017, it did so a year later. We would also be joined by Ishan, another friend who was interested in the project for its visual media, cultural and user interaction aspects.

Preparations involved elaborate discussions, implementation plans and tinkering. In a nutshell, we wanted to setup a reliable, off-grid knowledge server and a wireless network in the triangular shaped main valley using ‘mesh nodes’, and establish wireless links between at least three villages and a few schools far apart.

Our first major challenge — Power

Temperatures in Zanskar can range between +40C and -40C, and this makes it very hard to have reliable off-grid, battery powered systems. Our design for powering each mesh node involved:

  • a 10W solar panel,
  • four 18650 lithium-ion cells recovered from e-waste laptop batteries,
  • a battery protection module,
  • a MPPT solar charging module,
  • a boost module to provide a stable five volt output supply,
  • a domino core wireless ‘mesh’ module and
  • a wireless antenna made from a tin-can, and a satellite TV dish.

In my past experience with the domino core module in Spiti, I learnt that it can survive temperatures at least as low as -20C, and perhaps beyond that too. The real problem is managing the batteries. In the extreme cold of Zanskar, batteries would simply die if kept outside. Even inside homes, temperatures regularly fall well below freezing point. I learnt this the hard way during a winter in Spiti, as the aluminum bottle I used to drink water from was shredded apart like a piece of paper, when it was kept in one of the unheated rooms in the house.

To address this challenge, we divided each mesh node into two units. The first unit consisted of just the domino core and the antenna to be installed outside, as they must be kept as close to each other as possible. The second unit consisted of the remaining electronics and batteries, and was kept indoors in the room in which people live in during winters. A solar panel was hooked to this setup, and we had a solution that would work off-grid, hopefully survive cold temperatures, and would only require the occasional cleaning of the solar panel as basic maintenance.

The knowledge server

The next piece of the puzzle involved a server that would sit at the heart of the wifi mesh network, and provide essential services like VoIP telephony, access to content, and allow for the management of the mesh network.

To address this need, we used an Intel NUC mini-PC, and installed the Internet-in-a-Box (IIAB) software on it. IIAB started as a fork of the One Laptop Per Child’s school-server (XS) project in 2012. It aims to provide libre-licenced content and applications to remote communities worldwide — mostly in offline settings. IIAB deployments exist around the world — India, Haiti, Mexico, Ghana, Myanmar and many other locations. Also, since the IIAB is an open-source project running on the Linux operating system, it allows for unlimited customization — an excellent fit for our needs.

We built a custom power bank made out of sixteen e-waste batteries which would provide adequate backup to the server at nighttime and during periods of limited sunlight.

Wireless mesh

Even though there is hardly any internet in Zanskar, there is no shortage of smartphones. Application and content are shared between users through a sneakernet, usually over rounds of chai and hot water. In theory, the area is covered by a 2G network, but in reality, just one mobile tower serves the entire valley, and provides very unstable, unusable service. We thought that having a locally owned and locally maintained mesh network might offer a reliable alternative, especially so in the months when road access between villages is blocked by many feet of snow. A community owned mesh might also come in handy during times of emergency.

Our plan was to deploy around six mesh nodes in three villages, connect them to the knowledge server, and allow for services like VoIP calling and video streaming. This was the trickiest part of the setup, as none of us had been to Zanskar before, and no idea of the terrain and weather related challenges we might face.

The meshnode itself was made from a GLiNet domino core module, and loaded with a customized version of the openSECN mesh firmware (based on the popular open source openWRT project).

A wifi mesh node and a tin-can antenna

Content and media

Having deployed a mesh network in Spiti and Ladakh before, we had a decent collection of content on the IIAB. Before we went, Sandup, another friend and native of Zanskar visited many of the valley’s schools and interviewed teachers asking for what content they would like to see on their network. He gave valuable feedback, that helped us improve the selection of content available. Thanks to Prasenjit, we also had access to a large collection of teachings given by His Holiness the 14th Dalai Lama. However, as time before our trip was extremely short, we just about had the time to copy them into an external hard disk before leaving. How it would appear to users on a webpage was a big unknown.

What was also quite uncertain was how users would be able to upload, discover and share local media or content they would create in the valley. I had limited experience in this regard, and our very limited knowledge of where we were going did not help. Frankly, this aspect of our project made me the most nervous.

Travel & Logistics

Zanskar has been called, among other things, ‘the last place on earth’. Road access to Zanskar first involves getting to Leh, which in itself is quite an adventure. The road from Dharamsala (our base), involves crossing four 15000ft+ high passes to get to Leh itself, and then another equally long road trip crossing more high passes. There is a shorter route, but that would involve us trekking across a very high pass on foot, and we were very unsure of its viability. It didn’t help that we were carrying over 100kg of equipment with us, having already sent over 40kg in mail.

Road access to Zanskar is only possible during the summer-monsoon season. For the rest of the year, the only road into the valley remains frozen. There is a helicopter service that operates but is exorbitantly priced. The traditional winter route between Zanskar and Leh involves walking on foot on top of a frozen river — commonly known as a chadar trek.

In short, it was going to be nothing short of an adventure just getting to the place where we hoped to deploy this network. There was a running joke between us that we were headed to la-la-land (-la means a mountain pass in tibetan)

****

Soon the time to embark upon this adventure would arrive. The months leading up to this project were filled with lots of nervous excitement. Here was a chance to try out a new vision for the web in one of the remotest corners of the world, one as yet uncorrupted by the flaws of the internet — addictive social media, mass surveillance, advertisements, centralization to name a few.

Would our preparations and planning be up to the mark? Would our ideas of relevance and usefulness hold water? Would we and our equipment be able to make it to Zanskar in one piece? What will we learn from this experience?

Stay tuned for part 2…