All posts by Sridhar Dhanapalan

http://www.dhanapalan.com/blog/about/

Creating an Education Programme

OLPC Australia had a strong presence at linux.conf.au 2012 in Ballarat, two weeks ago.

I gave a talk in the main keynote room about our educational programme, in which I explained our mission and how we intend to achieve it.

Even if you saw my talk at OSDC 2011, I recommend that you watch this one. It is much improved and contains new and updated material. The YouTube version is above, but a higher quality version is available for download from Linux Australia.

The references for this talk are on our development wiki.

Here’s a better version of the video I played near the beginning of my talk:

I should start by pointing out that OLPC is by no means a niche or minor project. XO laptops are in the hands of 8000 children in Australia, across 130 remote communities. Around the world, over 2.5 million children, across nearly 50 countries, have an XO.

Investment in our Children’s Future

The key point of my talk is that OLPC Australia have a comprehensive education programme that highly values teacher empowerment and community engagement.

The investment to provide a connected learning device to every one of the 300 000 children in remote Australia is less than 0.1% of the annual education and connectivity budgets.

For low socio-economic status schools, the cost is only $80 AUD per child. Sponsorships, primarily from corporates, allow us to subsidise most of the expense (you too can donate to make a difference). Also keep in mind that this is a total cost of ownership, covering the essentials like teacher training, support and spare parts, as well as the XO and charging rack.

While our principal focus is on remote, low socio-economic status schools, our programme is available to any school in Australia. Yes, that means schools in the cities as well. The investment for non-subsidised schools to join the same programme is only $380 AUD per child.

Comprehensive Education Programme

We have a responsibility to invest in our children’s education — it is not just another market. As a not-for-profit, we have the freedom and the desire to make this happen. We have no interest in vendor lock-in; building sustainability is an essential part of our mission. We have no incentive to build a dependency on us, and every incentive to ensure that schools and communities can help themselves and each other.

We only provide XOs to teachers who have been sufficiently enabled. Their training prepares them to constructively use XOs in their lessons, and is formally recognised as part of their professional development. Beyond the minimum 15-hour XO-certified course, a teacher may choose to undergo a further 5-10 hours to earn XO-expert status. This prepares them to be able to train other teachers, using OLPC Australia resources. Again, we are reducing dependency on us.

OLPC Australia certifications
Certifications

Training is conducted online, after the teacher signs up to our programme and they receive their XO. This scales well to let us effectively train many teachers spread across the country. Participants in our programme are encouraged to participate in our online community to share resources and assist one another.

OLPC Australia online training process
Online training process

We also want to recognise and encourage children who have shown enthusiasm and aptitude, with our XO-champion and XO-mechanic certifications. Not only does this promote sustainability in the school and give invaluable skills to the child, it reinforces our core principle of Child Ownership. Teacher aides, parents, elders and other non-teacher adults have the XO-basics (formerly known as XO-local) course designed for them. We want the child’s learning experience to extend to the home environment and beyond, and not be constrained by the walls of the classroom.

There’s a reason why I’m wearing a t-shirt that says “No, I won’t fix your computer.” We’re on a mission to develop a programme that is self-sustaining. We’ve set high goals for ourselves, and we are determined to meet them. We won’t get there overnight, but we’re well on our way. Sustainability is about respect. We are taking the time to show them the ropes, helping them to own it, and developing our technology to make it easy. We fundamentally disagree with the attitude that ordinary people are not capable enough to take control of their own futures. Vendor lock-in is completely contradictory to our mission. Our schools are not just consumers; they are producers too.

As explained by Jonathan Nalder (a highly recommended read!), there are two primary notions guiding our programme. The first is that the nominal $80 investment per child is just enough for a school to take the programme seriously and make them a stakeholder, greatly improving the chances for success. The second is that this is a schools-centric programme, driven from grassroots demand rather than being a regime imposed from above. Schools that participate genuinely want the programme to succeed.

OLPC Australia programme cycle
Programme cycle

Technology as an Enabler

Enabling this educational programme is the clever development and use of technology. That’s where I (as Engineering Manager at OLPC Australia) come in. For technology to be truly intrinsic to education, there must be no specialist expertise required. Teachers aren’t IT professionals, and nor should they be expected to be. In short, we are using computers to teach, not teaching computers.

The key principles of the Engineering Department are:

  • Technology is an integral and seamless part of the learning experience – the pen and paper of the 21st century.
  • To eliminate dependence on technical expertise, through the development and deployment of sustainable technologies.
  • Empowering children to be content producers and collaborators, not just content consumers.
  • Open platform to allow learning from mistakes… and easy recovery.

OLPC have done a marvellous job in their design of the XO laptop, giving us a fantastic platform to build upon. I think that our engineering projects in Australia have been quite innovative in helping to cover the ‘last mile’ to the school. One thing I’m especially proud of is our instance on openness. We turn traditional systems administration practice on its head to completely empower the end-user. Technology that is deployed in corporate or educational settings is typically locked down to make administration and support easier. This takes control completely away from the end-user. They are severely limited on what they can do, and if something doesn’t work as they expect then they are totally at the mercy of the admins to fix it.

In an educational setting this is disastrous — it severely limits what our children can learn. We learn most from our mistakes, so let’s provide an environment in which children are able to safely make mistakes and recover from them. The software is quite resistant to failure, both at the technical level (being based on Fedora Linux) and at the user interface level (Sugar). If all goes wrong, reinstalling the operating system and restoring a journal (Sugar user files) backup is a trivial endeavour. The XO hardware is also renowned for its ruggedness and repairability. Less well-known are the amazing diagnostics tools, providing quick and easy indication that a component should be repaired/replaced. We provide a completely unlocked environment, with full access to the root user and the firmware. Some may call that dangerous, but I call that empowerment. If a child starts hacking on an XO, we want to hire that kid 🙂

Evaluation

My talk features the case study of Doomadgee State School, in far-north Queensland. Doomadgee have very enthusiastically taken on board the OLPC Australia programme. Every one of the 350 children aged 4-14 have been issued with an XO, as part of a comprehensive professional development and support programme. Since commencing in late 2010, the percentage of Year 3 pupils at or above national minimum standards in numeracy has leapt from 31% in 2010 to 95% in 2011. Other scores have also increased. Think what you may about NAPLAN, but nevertheless that is a staggering improvement.

In federal parliament, Robert Oakeshott MP has been very supportive of our mission:

Most importantly of all, quite simply, One Laptop per Child Australia delivers results in learning from the 5,000 students already engaged, showing impressive improvements in closing the gap generally and lifting access and participation rates in particular.

We are also engaged in longitudinal research, working closely with respected researchers to have a comprehensive evaluation of our programme. We will release more information on this as the evaluation process matures.

Join our mission

Schools can register their interest in our programme on our Education site.

Our Prospectus provides a high-level overview.

For a detailed analysis, see our Policy Document.

If you would like to get involved in our technical development, visit our development site.

Credits

Many thanks to Tracy Richardson (Education Manager) for some of the information and graphics used in this article.

OLPC Australia XO-AU OS 12 beta 1

The OLPC Australia XO-AU OS 12 has reached beta 1. This is based on OLPC OS 11.3.1 and Dextrose 3.

We’d really appreciate some testing. Please direct your feedback to the OLPC Australia mailing list.

Here is the notice I sent out to teachers:

———- Forwarded message ———-
From: Sridhar Dhanapalan
Date: 24 December 2011
Subject: Taking part in improving new XO software

Friends,

The 2012 OLPC Australia operating system, XO-AU OS 12, has reached a
beta stage of development. It has many improvements, and we looking
for feedback on how it works to help us create the final product. This
beta is suitable for testing, documentation and developing lesson
plans.

In early February, we will have a near-final release candidate,
suitable for trialling in classrooms. We are looking for clever
teachers to provide us with real-world feedback on how the software
works with their classes.

This is an opportunity for you to take part in XO development and
ensure that the device suits the needs of your classroom. We would be
especially interested to know how the connectivity and collaboration
works on your school’s networks.

To get started, visit our release notes page:
https://dev.laptop.org.au/projects/xo-au/wiki/120_release_notes

This page outlines the main changes in the new operating system. Go to
the “Beta 1” part of the Installation section. Installing the beta is
no different from installing the XO-AU USB 3 stable release: extract
the zip file to a USB stick and you’re ready to go.

To provide feedback, join our technical mailing list:
https://dev.laptop.org.au/projects/general/wiki/Technical_mailing_list

Following this, you can send your comments or ask questions at
olpc-au at lists dot laptop dot org

The OLPC Australia Engineering team are active participants on this
list, and we will reply. Remember, the better you can help us with
quality information, the better we can make the product for you 🙂

Regards,
Sridhar

Are you sick of highly paid teachers???

This has been making the rounds lately and is an absolute gem:

Teachers’ hefty salaries are driving up taxes, and they only work 9 or 10 months a year! It’s time we put things in perspective and pay them for what they do — babysit! We can get that for minimum wage. That’s right. Let’s give them $3.00 an hour and only the hours they worked; not any of that silly planning time, or any time they spend before or after school. That would be $19.50 a day (7:45 to 3:00 PM with 45 min off for lunch and planning that equals 6 1/2 hours). Each parent should pay $19.50 a day for these teachers to baby-sit their children. Now how many students do they teach in a day… maybe 30? So that’s $19.50 x 30 = $585.00 a day.

However, remember they only work 180 days a year!!! I am not going to pay them for any holidays. LET’S SEE…. That’s $585 X 180 = $105,300 per year. (Hold on! My calculator needs new batteries.) What about those special education teachers and the ones with Master’s degrees or higher duties? Well, we could pay them minimum wage ($7.75), and just to be fair, round it off to $8.00 an hour (but we shouldn’t get carried away). That would be $8 X 6 1/2 hours X 30 children X 180 days = $280,800 per year. Wait a minute — there’s something wrong here! There sure is!

The average teacher’s salary (nation wide) is $50,000. $50,000/180 days = $277.77/per day/30 students=$9.25/6.5 hours = $1.42 per hour per student — a very inexpensive baby-sitter and they even EDUCATE your kids!) WHAT A DEAL!!!!

Make a teacher smile; re-post this to show appreciation for all educators.

I don’t think the dollar values are for Australia — our minimum wage is higher than $3.00. The point should be obvious nonetheless: we seriously undervalue the people who are responsible for educating our children.

OLPC Australia talk at OSDC 2011

Update: my talk has been covered by OLPC News.

Here’s the video of the talk I said I’d be giving at OSDC 2011, titled Australia’s Toughest Linux Deploy­ment:

In it, I outline our educational programme and how the technology fits into it. Some key points:

  • we have a better version on YouTube of the video I show in the talk
  • we maintain a Policy Document, which provides an overview of our overall programme
  • OLPC Australia have two core principles in addition to OLPC’s original five
  • we have some support in government at different levels — for example, we were praised in federal parliament and the print media (paywall) by a prominent federal Member of Parliament
  • we have deployments across remote Australia — a feat that can only be managed through building self-sufficiency
  • our programme is showing beneficial results, and we are engaged in longitudinal and detailed evaluation
  • we have a comprehensive educational programme, with online training and certifications (such as our XO-cert course)
  • we are breaking dependence on special expertise and infrastructure — building sustainability and grass-roots support is key
  • deployments are made at the classroom level, which is more manageable than saturating a whole school at once
  • we don’t provide XOs without training — a teacher must earn a certification before they can receive XOs for their class
  • our support is focused on enabling schools and communities to help themselves, and each other
  • we have innovated in the technology space, with offerings such as the XO-AU OS, XO-AU USB, XOP and XS-AU
  • contextualising learning, for example through localisation, is a powerful tool to improve engagement from the child, school and community
  • we invite people to join our development efforts
  • there’s a nice surprise mentioned towards the end, which I shall elaborate upon in the near future 🙂

For those of you who have seen me speak about OLPC Australia at SLUG, this is a much more polished talk.

Speaking at OSDC 2011 on OLPC Australia

I am speaking next Thursday at the Open Source Developers’ Conference 2011 in Canberra. The title is Australia’s Toughest Linux Deployment. Yes it’s a play on the ruggedness and flexibility of the XO’s design to meet the needs of remote communities.

Here’s the talk abstract:

A 300,000 seat Linux deployment is nothing to sneeze at. What if those seats were actually children’s laps? By providing a flexible learning platform, OLPC Australia aims to create a sustainable and comprehensive programme to enhance opportunities for every child in remote Australia. What’s more, we plan to achieve this by 2014.

In focusing on the most remote areas of the continent, the mission is by no means easy. These areas are typically not economically viable for a business to service, hence the need for a not-for-profit in the space. Expertise for hardware and software is virtually non-existent. Settlements are small and spread very far apart. Environmental conditions, cultures and lifestyles vary wildly. They are very different worlds from the coastal cities where the bureaucracies are based.

Even within communities, differences abound. Schools often stand in stark contrast to their surrounds. Government and business interests have also made their marks.

This talk will outline how OLPC Australia has developed a solution to suit Australian scenarios. Comparisons and contrasts will be made with other “computers in schools” programmes, OLPC deployments around the world and corporate IT projects.

For example, standard sysadmin practice typically mandates tight, centralised control over all systems and infrastructure. The OLPC Australia approach is the exact opposite. By promoting flexibility and ease of use, the programme can achieve sustainability by enabling management at the grass-roots level. The XO laptops themselves are built especially for education. They are extraordinarily rugged as well as being inexpensive. They are also totally repairable in the field, with minimal skill required. Training is conducted online, and an online community allows participants nationwide to share resources.

Key to the ongoing success of the programme is active engagement with all stakeholders, and a recognition of the total cost of ownership over a five-year life cycle.

 

XO-AU OS 10.1.3-au3 and XO-AU USB 3 released

I’m pleased to announce two important software releases from OLPC Australia.

XO-AU OS 10.1.3-au3 is the latest iteration of our operating system for XO-1.5 hardware. It continues our process of refinement on top of OLPC OS 10.1.3, to better suit Australian educational environments. Details are available in the release notes.

I’ve also put together a summary of the improvements we have made to our operating systems in the past year.

XO-AU USB is a software ‘Swiss Army knife’ for XOs. It provides a boot menu to make important XO operations easily accessible, without the need to type any commands. It is our official means of installing our operating systems. Version 3 contains XO-AU OS 10.1.3-au3.

Cannes Lions award for OLPC Australia

OLPC Australia have been awarded a Bronze Lion at this year’s Cannes Lions International Advertising Festival, the advertising counterpart of the Cannes Film Festival.

I think this is fantastic recognition for a Free Software project, especially one that is focused on assisting children in some of the most remote parts of the world. I feel honoured to have been part of this success.

We’re happy for people to get involved to help us in our mission. If you’d like to participate, especially (for me) in the technical field, please get in touch with me or contact OLPC Australia through our Web site.

HTML5 in Sugar

In my last blog post, I made the suggestion that Sugar integrate HTML5 more closely to allow for the creation of activities in standard Web technologies. The Karma Project has since been pointed out to me, and the demos look impressive. Unfortunately, its progress looks to have stalled. There is now consideration happening in the community about moving Browse to a WebKit-based alternative, possibly Surf.

It seems like now is the time to revisit the notion of integrating HTML5 into Sugar itself. I feel that this can achieve a far more powerful outcome than just swapping Browse with Surf. The primary weaknesses of HTML5, its immaturity and dearth of good development tools, are being addressed. Microsoft and Adobe are continue to move towards HTML5, which can only be a good thing.

We have the chance to tap into the current rush of developers creating Web applications. We don’t need to (and can’t afford to) go to the extreme always-online level of Chrome OS, but I think the developments in that space are really showing what HTML5 can do in terms of applications development. Take the Chrome version of Angry Birds, for example. Written (almost) entirely in HTML5/JS (I think the “almost” part could have been implemented in HTML5 as well), it’s a fantastic example of what can be achieved. More than a mindless game, the physics engine is realistic enough to become a fun educational tool. It’s so much fun that most kids won’t even realise that they’re learning.

Why ‘Free and Open’ matters

Adobe is dropping Linux support for their Adobe AIR development platform. To be honest, I don’t really care. Why? Because I’ve been careful enough to not tie my efforts to a proprietary platform.

I’ve had several groups offer to write applications/activities for OLPC Australia using proprietary tools like AIR. I’ve discouraged them every time. Had we gone with the ‘convenient’ route and acquiesced, we would have been in quite a spot of bother right now. My precious resources would have to be spent on porting or rewriting all of that work, or just leaving it to bit-rot.

A beauty of Sugar and Linux is that they are not dependent on a single entity. We can develop with the confidence of knowing that our code will continue to work, or at least can be made to continue to work in the face of underlying platform changes. This embodies our Core Principle #5, Free and Open.

Free and Open means that children can be content creators. The television age relegated children (and everyone, for that matter) to just being consumers of content. I have very fond childhood memories of attempts to counter that, but those efforts pale in comparison to the possibilities afforded to us today by modern digital technologies. We now have the opportunity to properly enable children to be in charge of their learning. Education becomes active, not passive. There’s a reason why we refer to Sugar applications as activities.

Growing up in the 80s, my recollections are of a dynamic computing market. Machines like the ZX Spectrum and the early Commodore models inspired a generation of kids into learning about how computers work. By extension, that sparked interest in the sciences: mathematics, physics, engineering, etc.. Those machines were affordable and quite open to the tinkerer. My first computer (which from vague recollection was a Dick Smith VZ200) had only a BASIC interpreter and 4k of memory. We didn’t purchase the optional tape drive, so I had to type my programs in manually from the supplied book. Along the way, I taught myself how to make my own customisations to the code. I didn’t need to learn that skill, but I choose to take the opportunity presented to me.

Likewise, I remember (and still have in my possession, sadly without the machine) the detailed technical binders supplied with my IBM PC. I think I recognised early on that I was more interested in software, because I didn’t spend as much time on the supplied hardware schematics and documentation. However, the option was there, and I could have made the choice to get more into hardware.

Those experiences were very defining parts of my life, helping to shape me into the Free Software, open standards loving person I am. Being able to get involved in technical development, at whatever level of my choosing, is something I was able to experience from a very early age. I was able to be active, not just consume. As I have written about before, even the king of proprietary software and vendor lock-in himself, Bill Gates, has acknowledged a similar experience as a tipping point in his life.

With this in mind, I worry about the superficial solutions being promoted in the education space. A recent article on the BBC’s Click laments that children are becoming “digitally illiterate”. Most of the solutions proposed in the article (and attached video) are highly proprietary, being based on platforms such as Microsoft’s Windows and Xbox. The lone standout appears to be the wonderful-looking Raspberry Pi device, which is based on Linux and Free Software.

It is disappointing that the same organisation that had the foresight to give us the BBC Computer Literacy Project (with the BBC Micro as its centrepiece) now appears to have disregarded a key benefit of that programme. By providing the most advanced BASIC interpreter of the time, the BBC Micro was well suited to education. Sophisticated applications could be written in an interpreted language that could be inspected and modified by anyone.

Code is like any other form of work, whether it be a document, artwork, music or something else. From a personal perspective, I want to be able to access (read and modify) my work at any time. From an ethical perspective, we owe it to our children to ensure that they continue to have this right. From a societal perspective, we need to ensure that our culture can persevere through the ages. I have previously demonstrated how digital preservation can dramatically reduce the longevity of information, comparing a still-legible thousand-year-old book against its ‘modern’ laserdisc counterpart that became virtually undecipherable after only sixteen years. I have also explained how this problem presents a real and present danger to the freedoms (at least in democratic countries) that we take for granted.

Back in the world of code, at least, things are looking up. The Internet is heading towards HTML5/JavaScript, and even Microsoft and Adobe are following suit. This raises some interesting considerations for Sugar. Maybe we need to be thinking of writing educational activities in HTML5, like those at tinygames? Going even further, perhaps we should be thinking about integrating HTML5 more closely into the Sugar framework?

I’ll finish with a snippet from a speech given by US President Obama in March (thanks to Greg DeKoenigsberg for bringing it to the attention of the community):

We’re working to make sure every school has a 21st-century curriculum like you do. And in the same way that we invested in the science and research that led to the breakthroughs like the Internet, I’m calling for investments in educational technology that will help create digital tutors that are as effective as personal tutors, and educational software that’s as compelling as the best video game. I want you guys to be stuck on a video game that’s teaching you something other than just blowing something up.