Fixfest Data Event

Fault categorisation

We would like to be able to analyse common fault types for devices that we have seen at our community repair events.

However, the problem field, where the fault information is recorded, is currently free text. To analyse it, the faults need to be ‘coded’ – i.e. classified into common types.


In these sheets, read through the text provided in the problem field, and try to select the appropriate fault type from the dropdown.  In addition, if possible, try and categorise the problem into a fault category of either ‘Hardware’ or ‘Software’.

Category Publisher Language Number of records Link
Tablets Anstiftung German 30 Anstiftung – Tablets
Laptops Anstiftung German 170 Anstiftung – Laptops
Desktops Anstiftung German 25 Anstiftung – Desktops
Tablets Restart Mostly English 95 Restart – Tablets
Laptops Restart Mostly English 262 Restart – Laptops
Desktops Restart Mostly English 24 Restart – Desktops

Restart data is since March 2019 – previous data was categorised during and following Open Data Day.

Some questions…

  • What are the most common faults in each of the product categories?  In the whole cluster of tablets, laptops and desktops?
  • How do different faults correspond to repair outcomes?
  • Do we see more hardware or software problems?  How do they correspond to repair outcomes?
  • Do we see a similar split of of fault types between Anstiftung and Restart networks?

Raw data download

The full ORDS dataset is available here, if you would prefer to try and map the free text to fault classifications in a different way. (Note that it includes Restart data from before March 2019, which has already been classified – but is not currently included in the ORDS dataset).

An example R workbook from Open Data Day is available here.