Learning to read with One Tablet per Child

Can tablets make a difference to a child learning to read for the first time, without a teacher or traditional classroom structure? That’s the question we are exploring with our reading project, currently underway in Ethiopia.

A few dozen children in two rural villages have been given tablets which they are using for a few months. They are interested in learning to read English, and understand this is something they can learn with the tablets; which also come with hundreds of children’s apps.

They are equipped with software that logs all interactions, building up a clear picture of how each tablet is being used. Data from the tablets is gathered each week and sent back to the research team, which also rolls out new updates to the tablets week by week.

Richard is in Ethiopia this week, to get better first-hand knowledge of how the tablets and other infrastructure are holding up, and a visual sense of how they are being used.

“if a child can learn to read, they can read to learn”

Low-power solutions: energy harvesting and cultural implications

Last November, Richard Smith gave a talk on potential power sources for OLPC at the Energy Harvesting USA conference.

Ars Technica recently reviewed the XO-3 and XO-1.75 with an eye towards the future implications of low-power computing, discussing power generation by hand, by bike, and by water wheel as well as through solar panels.

Of course we’re not done with the low-power revolution; phones and computers – even the latest XOs – are still too power hungry to be quickly and easily charged by ambient light (as solar calculators are) or by hand (despite the simplicity of hand cranks, legs are much better suited for generating power than arms and hands). So while alternate charging works it requires explicit attention and preparation.

But in places without electrical infrastructure that have some steady source of power, computers and computing can increasingly be part of everyday life.

XO power data: power draws for plugged-in laptops

Guest post by Richard Smith

We’ve finished testing power consumption while plugged into a 230V ac wall outlet for the XO-1 and the XO-1.5. The new machine performs well while suspended, and suspends very smoothly.   The 1.5 charges faster and using less power to charge the same battery.  It also draws slightly more power when in high use, thanks to its variable CPU.

See the chart below, which includes the power draw of the AC adapter.  Battery-only numbers will be significantly lower, in particular for idle and suspend, but are a bit harder to measure cleanly.  There is no comparison chart for that yet.

Power draw at the wall (XO + adapter + backlight)

Scenario XO-1 XO-1.5
Full charge 56 Wh 47 Wh
Idle 8.5W 7.2W
High 9.5W 9.7W
Suspend 5.2W 2.85W
  • Full Charge: The amount of energy it takes to completely charge a dead battery, using an adapter (power needed for bulk charging of batteries may differ).
  • Idle: Laptop sitting at the Sugar home screen, with power management disabled, backlight on full, charging.
  • High: Laptop running the Record activity’s “preview” mode.  Power management disabled, backlight on full, charging.
  • Suspend: Laptop with power management enabled, suspended, charging.

Calculating low-power options:

  • The backlight draws close to 1W – you can shave that off of the idle and high numbers turning it off.
  • Most of the power draw on suspend is to the adapter – you could view that as an upper bound on how much to factor out of the other numbers for battery-only power usage.