Internet Connectivity: The Achilles Heel of the Peru Deployment


There have been many articles and reports written recently about the Peru deployment of XO laptops; I would like to try and supplement these criticisms with some first-hand experience that reveal the fundamental problems. I am Joshua Debner and this commentary is based on 3 months as an OLPC Intern in the field with the largest deployment of XO's in the world: OLPC Peru.

The Peru situation is frequently compared to Uruguay because of its similar geographic location, language, and size of deployment. Many people blame physical terrain for some of the major differences between the successes of the two deployments, but after traveling in rural Peru, I do not think that physical terrain is the main thing holding back the program's success.

Peru began purchasing the laptops in June 2007 after the success of Uruguay's program, but they failed to reach the same level of planning that Plan Ceibal required, and they missed some of the fundamental principles of OLPC in the process.

Current Situation

This is a video that I watched when I was first introduced to the OLPC program. It outlines the 5 principles that the every deployment must adhere to in order to be a part of the program. These principles are:

  1. Child Ownership
  2. Low Ages
  3. Saturation
  4. Connection
  5. Free and Open Source

The Peruvian Ministry of Education does not follow the 4th principle of OLPC deployment (connecting the laptops to the Internet); this creates numerous other problems affecting the software image and teacher training.

Root Problem: #4 Connection to the Internet

The OLPC program is often described as a way to open doors to children who don't have access to the same resources you and I do. You aren't just giving a child a laptop; you are giving him a window to a world full of knowledge and exploration. The problem is that the schools we worked with in Peru did not have Internet that was accessible to the students' laptops. [Our partner in crime, Jeff, worked in over 20 schools all without Internet].

Although a few of the towns we worked in did have an Internet connection available within the town, they did not have a wireless connection available for the XOs, nor a plan or idea of when or how they could make this happen. One of the schools we worked in had Internet connected to their XP desktops. When I asked if they had considered buying a wireless router so they could use Internet with the XO's, they had no idea this possibility existed or how much it would cost.

This is a problem that can be directly solved by the Peru Ministry of Education. If you don't install Internet in the town, don't deploy laptops there. Yes, they might not get the coverage that there currently is in Peru in terms of number of laptops, but that might allow them to have a budget for Internet at each school and create deployments that are actually *updatable* and connected.

It's about quality vs. quantity. This, for me, is the core fundamental problem with Peru's plan for deployment. The Internet-less deployment creates a domino effect of problems which overshadow the original source of the problem.

Software Problem: #5 Free and Open Source

The reason OLPC wants all their deployments to be running free and open source software is so that "the laptop itself can easily grow and adapt". This includes being able to update the software easily and regularly, because if you've ever written software you know that it is never done, especially open source software. There will always be bugs to fix, patches to add, and features to implement.

As a computer engineer, I can say that the custom Peru image of the Sugar software that was shipped on the laptops throughout the country is at best, embarrassing. It doesn't really hit you until you are trying to demonstrate and explain an activity to a teacher and it just fails to work. Then you explain that you didn't invent the laptops and they are people still fixing things to make it better but if should work most of the time in your classroom.

The problem is that the only method of updating this software, fixing any bugs or making any changes, is to manually flash all the laptops with a flash drive. Let me crunch some numbers for you.

The Laptop Flashing Problem by the Numbers

  • 300,000 laptops - the first batch of laptops
  • 100 - assumed number of people to go and flash these laptops around the country
  • 3,000 - total number of laptops/person
  • 5 minutes - assumed flash time, and each of these workers carries two flash drives.
  • 3000 / 2 * 5 = 7500 minutes - How long will each worker have to flash
  • 7,500 minutes - over 5 straight days of flashing 24/hours a day not including transportation.
  • 8 hour work day - 15 days (three weeks) of straight flashing. Ouch!

After crunching these numbers, you can see the importance of that Internet connection that they left out with principle #4. At the end of our internship, we were presented with XO's running the new Peru software image running the latest version of Sugar.

It was so exciting to see so many of the bugs we had been struggling with the past month fixed in this version, but it was saddening because of the low probability that all the XO's will receive this software update and the next version that will be released as soon as they begin updating machines to this version.

Teacher Problem

One of the most common complaints from people who have worked with deployments in Peru is that more teacher training is needed. In every school that we worked with, we had to start from scratch (the beginning, not the activity) because the teachers weren't confident with the simplest of tasks. The teachers felt alone and abandoned with no one guiding them or supporting them.

We had one successful encounter where one of the teachers of a school that we had already been to was contacted by a teacher we weren't visiting. The teacher we had taught was able to help teach this other teacher while reinforcing her own skills. The problem is that the only reason this happened is because the two teachers are best friends and call each other to keep themselves from getting bored in the rural towns they work in.

Imagine that instead of these teachers feeling helpless and alone, there was a network of teachers who were in similar circumstances learning the same things. This network could help the teachers bond together, share resources and even provide feedback to the Ministry on things that they are lacking (like thumb drives).

When we were back meeting with the Ministry of Education we were disappointed to see a very low confidence level in the teachers of Peru. When we asked about some specific problems with the software or training programs they were dismissed with the teachers wouldn't know how to use them anyway.

When we were working with the teachers on a 1-on-1 basis they were very excited to learn, and showed a strong ability to retain this information contrary to the Ministry's view.

Looking Ahead: 4 Steps to Stop the Bleeding

  1. Internet: Connect your current schools. Stop your current deployment plans and fix the ones that are already existing. Make the money that you've spent not go to waste before it is too late. Figure out a scalable plan for providing the schools in your country with Internet, whether you use satellite, 4G, or cell phone links, make a plan and begin to implement.

  2. Update: Push through this new software update currently said to be done by end of January. It is going to be tough to get it through and out on all the machines but it is necessary because this update includes support for over the air updates of activities and hopefully the software image. This will be critical once your Internet network is built up and will make your next update hopefully 6 months (instead of 2.5 years) later much less painful.

  3. Teacher Network: Everyone says teacher training is needed, but teaching individual teachers is a lost cause if they aren't connected and able to share their experiences. Instead work to build a network and tools for teachers to share lesson plans, or communicate with each other. Even putting up something as simple as a phone list of other teachers in the area where teachers can begin to network the old school way would be useful.

  4. Leverage the Intern Program: The final point is that you, the Peru Ministry of Education should leverage the model of the intern program set up by OLPC. There are hundreds of qualified competent students in the US and abroad who are willing to work for next to nothing in order to have an experience that will in some way change them. Our transportation wasn't covered and many of the towns we stayed in provided food for us, it was their way of telling us that these laptops are important to them and their town. Because they appreciated us trying to help them, our costs are minimal. Leveraging interns and students from within Peru would also be very beneficial if you can find interested candidates.

Note about the OLPC intern program

It is with great sadness that I inform you the OLPC intern project in its current form is now dead. OLPC would rather not spend money on field interns who are trying to make a difference at the ground level, but instead would like to send graduate students who have stipends from their schools to come and sit in a city office and write papers about what the government should do next.

The Peruvian Ministry needs to get at ground level with the teachers and students using the laptops and figure out needs for themselves. Stop writing papers and theorizing; Act.

Related Entries


These firsthand reports are invaluable to making changes and improving deployments. It's good process and feedback to the respective organizations and communities.

Networking or rather, dependable network infrastructure has always been a foundational requirement for the successful deployment of 1:1 programs worldwide. My company learned a lot about this firsthand when Maine adopted a state-wide program and we've also seen some of these same networking issues in Alaska.

I really applaud and admire the work of the OLPC interns and volunteer techs who have been trying to communicate back up the chain of command to the government agencies and sponsoring organizations. There is no substitute for first-hand experience in the classroom and on campuses.

Maybe it already exists but a network/Internet availability map or Google mash-up per region could be another opportunity for grants/funding to be targeted at broader infrastructure first. A dependable high speed network benefits a larger community of users; not just classrooms.

These are the growing pains of many 1:1 programs; not just OLPC.

Scott Love
Palo Alto, CA

This is a very good article and full of vital information. For years, I have been saying the same thing about infrastructure needs and high-quality, in-class professional development for teachers. It seems that these messages, and this article says much about those who insist on doing it "their way" and we will make the same mistakes that others do, but in a different setting.
Here is the message again: No infrastructure that works efficiently and effectively, no really good usage of laptops. No infrastructure that enhances the learning and teaching environment, no really good usage of laptops. No effective, productive professional development for educators that supports a higher quality integration of technology, no really good usage of laptops.
Whenever 1:1 deployments are presented, it seems to inevitably shift from a "Field of Dreams" approach to an "Arena of Educational Realities."
Listening to the practitioners (teachers and interns) state realities will simply make the chances of success increase.


Well said. I might also add that the state of Maine walks the talk with respect to professional development. It's integrated into the schools of education in the surrounding regional colleges/universities while also provided year round to teacher. In fact, Maine's PD provides credits and additional opportunity for teacher advancement. You can't have best practices without practice.

My own opinion is that technology itself doesn't deliver outcomes but it can definitely impact outcomes whenever you can provide access to digital content because it's more cost-effective than buying out-dated textbooks.

Call this the beyond the textbook movement. MIT's Scholar Series is a real-world example of what's possible today; world class instruction in physics for anyone with an Internet connection.

OLPC and 1:1 programs put in this context are absolutely going to impact education and outcomes. And it starts with network infrastructure and teacher training as a standard requirement for continued success. Just takes time and investment.


Together with my wife we are now preparing to get involved as a 'rambling deployers' in OLPC Asia area. At first we identified two major fields the project needs improvement in, one of which is maintenance & support. Obviously, the connectivity is a part of it, as well as of the other one - educator's education.
Just to let you know, I am now researching possible use of packet radio ( solutions to:
1. Set up links between XSs to extend mesh connectivity beyond single community area.
2. Provide an uplink for asymmetric satellite download (selective broadcast), making massive content delivery and software upgrade possible without expensive 2-way VSat terminals.
3. Establish long-distance remote administration (text-based) for XS and XO management for isolated networks.

All comments and contributions welcome, either here or directly to me. As soon as get any research conclusion, I will report back here.

Greetings from Warsaw, Poland.
Piotr / Petros

thanks for this article, Joshua.

I was in Peru as well during the last year, as researcher, and of course tried to help out where I could...
I was wondering about your recommendation to connect all the current schools with the internet, because this to me seems, well, quite impossible at this moment, because in many of the places (those I have visited in Puno and Madre de Dios) the only possibility to connect is via satelites and that is of course much to expensive, and I think they won't spend that amount of money for a little village-school with 10 or 20 students...when I talked to people at DIJETE I was told they are planning to connect all the schools with minimum 100 students, which will rather be in towns than those little abroad villages...
and actually I wonder how and what for they should use the internet...there is no "culture of exchange" for information, that is why there are so few teacher networks, because people are just starting to realize that it helps to share knowledge outside their direct networks...I think it will take some time before people are ready to really apreciate and use the internet and finally to me (having studied anthropology) it seems a very "western" idea, because we are so used to the internet and use it for everything, but also we have slowly learned and developed ways of using it and there were computers long before the internet...

I would really be interested in discussing this more...

The Peruvian Ministry of Education does not follow the 1th principle. The laptops are property of the schools, children are only the users not the owners.