Replacements needed Using MIL-STD-80 Notebook Computers
In this example, computers are initially delivered at a rate of approximately 1,750 per month (excluding replacements) until month 60, resulting in a total of 560,000 computers being deployed.
Replacements begin during the first month but increase until they reach a peak at approximately 4 years of age. Replacement computers are then added to those being deployed as the old ones require replacement.
After the 60-month initial deployment period, the rate of replacements decreases to around 150 per month, including replacements for replacements as necessary.
Replacements Needed Using Inexpensive Generic Notebooks
Due to the shorter lifespan of the computers, there is no decline in the number of replacements required after the initial deployment ends (at month 60 with the same initial numbers as above). Instead, the number of replacements needed continues to increase, reaching a point where approximately 54,000 replacements are necessary every month.
As a result, the implementation becomes stuck in a cycle where a significant proportion of all computers must be replaced each month, leading to an unsustainable system.
Replacements Needed under One Laptop Per Child Model
This model has been tried in Africa and elsewhere with “limited success” at best. It uses a similar probability distribution as the Inexpensive Generic Notebook model, but magnifies the problems because of the large number of computers needing replacement each year.
As a result, the amount need simply to keep up with broken machine replacements becomes unaffordable after only a few years and never gets better.
NOTE: The projections presented are based on the available information regarding the expected lifespan of the machines being compared. This information is presented in the form of probability distributions, which are derived from the information that is currently available. However, in order to make our projections more accurate, we will be electronically collecting additional information based on the machines that we have deployed in African centers. By gathering this data, we will be able to refine our probability distributions and make more accurate projections about the lifespan and replacement needs of the machines in the future.
The number of replacement laptops that will be required from a particular batch of deployed computers is predicted using the following graphs. These display the odds over a ten-year period.
For instance, there is a small chance that a new machine won’t function, but the majority of machines will need to be replaced a few years after deployment. These diagrams depict the likelihood that a machine will require replacement within a specific month.
Replacement Model for Rugged Laptops
This diagram shows the probability distribution for a rugged laptop.
Replacement Model for Inexpensive Laptops
This diagram shows the probability distribution for an inexpensive laptop. Notice that the average lifetime is not only shorter, but most of these laptops need replacement at the same time (the curve is not as wide and the edges are steeper than the one above.
Replacement Model for One Laptop Per Child Laptops
This probability distribution closely resembles the Inexpensive Laptop model.