• approximately 10 classrooms;
• 50 teachers;
• 3000 school-aged;
• no electricity; and as of today
• 3 computers;
• 1 high-speed internet connection.

Today was the culmination of a couple of days of negotiations and waiting (see previous post) to get a small computer center set up at Bagrami school, a couple of hundred meters from the Taj guest house. The school does not have access to city power and does all their teaching on black boards in class rooms stuffed wall to wall with school desks. Because there are so many kids around, teaching is done in shifts with the girls attending in the mornings and the boys in the afternoons.

The school now also has 3 Inveneo computers running off a motorcycle battery plus 2 Linksys routers running of the same battery to provide them with internet access. The Inveneos were designed to run off 12V DC, making a car or motorcycle battery the ideal power source. The Linksys routers also conveniently run off 12V DC. We’ve been thinking more and more of just standardising all our hardware on 12V to make life easier in areas not served with grid electricity.

The Jalalabad fabfi network on 3 September 2009

The image above makes the geek in me very excited. Smari just explained it to the rest of us and I just colour coded his explanation. The relevant bit for this post is the blue connection pair coming out of the red source (the satellite internet connection at the Taj). The blue pair shows the connection from the roof of the Taj to Bagrami school. They currently have their computer switched off, otherwise you would have seen another link on the end of the blue chain. The second red node is the large reflector on the water tower in the center of Jalalabad. We set that up in January as a relay point for other sites in the city. The green pair shows the connection to the hospital right next to the water tower. Note that this image was pulled from the fabfi site where it is updated in real time to show the current extent of the Jalalabad network. The ability to generate these dynamic statistics is part of the offering of Fabfi 2.0. As we switch more links from Fabfi 1 to Fabfi 2, you should see more links appearing on the graph.

So, geekiness aside for a moment, a school that used to be completely off the grid now has access to computers with basic software installed. (OpenOffice and the web browser seem to be the most popular for now.) Not only can they learn and improve their knowledge of computers, highly valued in Afghanistan, but they can connect with the rest of the world in order to pull in ever more knowledge relevant to them and push out information about their existence and their place in the social fabric of the web.

]]> http://blogs.fabfolk.com/carl/?feed=rss2&p=24 0 http://blogs.fabfolk.com/carl/?p=12 http://blogs.fabfolk.com/carl/?p=12#comments Tue, 01 Sep 2009 09:57:47 +0000 Carl http://www.srcf.ucam.org/~cs482/blog/?p=3 Waiting is a large part of the game when on a tech mission in Jalalabad. There is the waiting for transport where finding an available car and driver could be something that happens at 5am, 1pm, or never. Probably our most helpful and responsive local driver is Hajji jan. This is of course not his real name, but an indication that he has done his pilgrimage to Mecca. Hajji is amazing in that he will be on time any time of the day and is perfectly willing to ferry a bunch of crazy white people around town.

There is waiting for the sun. Jalalabad is a hot place in the summer with temperatures often in the 40s (centigrade). This makes it very difficult to do anything other than hang out in an air-conditioned room for the hours around mid day. Yesterday was an exception though. We had a huge thunder storm in the middle of the night. It reminded me of a tropical storm from a Hollywood movie: high winds, heavy rain and constant atmospheric lightning. The bad news was that a lot of our kit was sitting just outside on tables and on the ground–ostensibly under a bit of cover, but nothing that could withstand this new weather phenomenon. It took 4 or 5 trips of running outside in my shorts, grabbing armfuls of computer equipment (cases, screens, wireless routers, batteries) and jamming the doorhandle with my elbow to get back inside. I was a bit of a drenched rat afterwards, but pretty high on adrenaline. The next morning and day were just gorgeous–the sky was clear of dust and pollution, the air was cool and you could see all the way to the second set of mountains to the north. We even had most of our computer equipment survive.

There is the waiting to get healthy. Last week’s bout of what could only have been H1N1 nearly took me out on the first day and ended up consuming about 5 days of my life. Day 1 was one long struggle through delirium from dehydration and being too out of it to move the 5cm to the nearest water bottle. I remember little of that day other than heat and thirst. Day 2 saw me able to move but still very weak physically. Walking was a chore and balance and coordination non-existent. My body was clearly still fighting off the infection; resting heart rate: 120bpm. From here on it was a slow but sure recovery. Day 3 gave me an extremely sore throat and some small return of physical strength. Resting heart rate: sub 90bpm (although any physical activity would bump it up to 120bpm again). Days 4 and 5 was dominated by coughing spells with my body finally having overcome the virus and in the process of cleaning itself of remaining toxins. Resting heart rate: sub 70bpm.

Right now, Smari and I are waiting for Naqibullah to arrive at the Taj so that we can walk over to Bagrami school. We’re supposed to meet ‘at 2pm’, which means some time between 2pm and 3pm… probably. Currently, it is 2:25pm. The plan is to have the 2 Inveneo computers up and running along with a Fabfi 2 router pointed at the Taj for internet access. With all of this set up the teachers there will be able to use basic applications like a word processor and spreadsheet (OpenOffice on an Ubuntu Linux box) and access the internet for as long as their 12V motorcycle battery lasts. Pretty cool.

The not-so-ready fablab site

The advanced team for the Fab5 conference has been in India for 2½ weeks now. The first 2 weeks was the time for local reconnaissance, installing a FabLab for the conference and preparing FabFi 2.0, the new generation of FabLab made antennas for wireless networking. We decided to get bicycles on the 3rd day or so and have been having a lot of fun cycling through the ridiculous Indian traffic. There aren’t really any enforced rules–you just kind of drive/cycle/walk where you need to and other traffic kind of flows around you. It’s terrifying in the beginning, but very efficient once you get used to it. The smaller your vehicle, the easier it is to squeeze yourself through the sometimes marginal gaps between trucks and buses. This makes bicycles and motorcycles (very popular) the best way to get around.

Kenny and Nadya testing the ShopBot

We spent the first two weeks getting the lab set up and generally preparing for the conference that’s currently running. This wasn’t without obstacles as the site was still under constructing until a few days before the start of the conference. We had to unpack and install our equipment around people laying floors, installing air conditioners, building staircases and painting the walls. Also, about 2 days after our arrival in Pune, the first person in the city died from H1N1 (the death toll is roughly 13 people by now, with around 1200 reported infections). The local media have been spectacularly sensational in its reporting and schools (including the college where we installed the lab and are having the conference) were closed for the entire second week of our install time–with the consequence that we had no student help in that time.

FabFi

One of the big topics of this year’s conference is building wireless internet infrastructure using a FabLab. Keith Berkoben (a technology policy grad student at Harvard) has been working on this as a hobby for some time now and we’ve frantically been building and installing antennas over the past week and a half. For my part, I’ve been learning more from Keith about router setup and configuration as well as going through the motions of hauling antennas up tall structures, pointing them, tying them down, and creating and testing the radio connections.

The antenna on the hill

One site of particular interest is Vigyan Ashram near Pabal, India. The FabLab concept was born when the founder of the ashram visited MIT in Boston and espoused the virtues of solving real problems locally in under served parts of the world. As a demo for the conference delegates we set up a multi-hop link from the room where the first FabLab ever was housed (it’s since moved next door to a larger room), to the nearby radio tower, to a hill at the other end of the ashram. We ran a fairly high bandwidth video conference over this link and had perfect audio and video all the way. Very exciting stuff.

Amy over the video link

Where we’d like to see this go next is Soshanguve, South Africa. The Soshanguve FabLab has been around since 2006 and has been building strong relationships with the local community, including schools and colleges. They would now like to use this social network to build a wide area computer network. My plan is to help them build the first wireless link by advising remotely, and by visiting Sosh when I’m in South Africa in September.

The lab at Sosh has a fairly poor internet connection and the schools to which they’d like to connect typically have no connection. Apparently many of the schools do have underutilised computer labs though. The Sosh folks were thinking of simply giving the schools slow internet access through their existing Telkom uplink. I’ve been convincing them to do something else instead. The plan is to put a server in the Sosh lab and to provide educational content to the schools/colleges directly from the server. Some obvious candidate content is Wikipedia, MIT OpenCourseWare, the Shuttleworth Foundation’s Free High School Science Texts project. Since this content is served locally (and not over a Telkom line) it is high bandwidth and there is virtually zero running cost. We would also use the server to cache WWW pages locally so that frequently visited sites run much faster. Eventually it would be useful to put a fat internet pipe at Sosh (at non-trivial cost) and provide good internet access to the rest of the FabFi network. This is not necessary for the first iteration though.

This model of serving high bandwidth static content locally at almost no cost, and serving dynamic lower bandwidth content through a shared internet uplink has already been implemented in Jalalabad during our visit there in January. We’ll be heading back to Jalalabad for two weeks, starting next week to roll out the current FabFi version 2 (they’re using version 1 right now) and to do some work in the FabLab there.

Possibly the greatest thing that happened in Afghanistan is that the locals starting building more antennas after the FabFi team left. This is the direction in which we will point Sosh also. We help build the first link or two, show them how to do it in the process, and rely on local users to build more wireless links–the more the better. This would allow them to provide access to information to the sites that matter to them.