Over half a million records of open data on repair

October 16, 2026

For Repair Day 2026, we’re happy to release the latest Open Repair Alliance dataset – our vast open dataset of information on items and repairs seen at community repair events around the world.

Just like the community repair movement itself, the volume and value of the data has grown again in 2026. With both EEE and non-EEE data included this year, there’s more data than ever available for download and analysis.

600,000 records of repair

The ORA dataset now represents a remarkable log of over half a million records of repair attempts from the community repair movement. Every single row of data represents multiple people working together in their community to fix their broken things.

Since we’ve started collecting this data, there are now:

• 619,875 repair attempts logged
• From 44,252 events
• Held by 1,728 repair groups
• In 34 countries worldwide

Looking only at the most recent 12 months, between 1st August 2025 and 31st July 2026, we’ve seen:

• 154,464 repair attempts, logged by 1,023 groups, at 12,821 events, in 20 countries
• That’s an average of nearly 13,000 items logged per month
• On average, over 400 repair attempts each day!

And that’s just the data logged – we know that there are many more groups running events worldwide, not yet represented in this data.

And looking purely at our ‘traditional’ dataset of electrical and electronics equipment repairs performed at community events, the growth of the dataset is again remarkable – we went from 305,649 in last year’s release, to 416,284 this year, an increase of 36% in just one year.

Download the data

Where does the data come from?

The data in the Open Repair Alliance dataset is collected by volunteers in community repair organisations all over the world. People run community events like Repair Cafes, Fixit Clinics and Restart Parties, where they help people from their community fix their broken items.

Many groups then log the data about the items they saw at their event, digitally on an online platform (more on this here). That’s where we come in – the raw data for the Open Repair Alliance dataset is provided to us from these various sources, and we do the work of combining these together into one dataset.

New sources of data

The data is truly global, and this year we’re very happy to see two new countries represented – Malaysia and Liechtenstein . And we’re equally excited to have added a new data source: iFixit’s Community Repair Organizer (community.ifixit.com). The Community Repair Organizer is powered by an installation of The Restart Project’s Restarters.net software and is the place for repair groups in the United States to manage their activities.

We’re always looking for repair data from new sources – so if you are collecting data from a country not yet represented, or you represent a network or platform that collects data from multiple groups and would like to contribute your network’s data – get in touch. If you’re a repair group and you want your data to be part of the Open Repair Alliance data, why not start adding it to one of the available platforms?

Cleaning up brand data

Beyond just collecting data, we aim to make the Open Repair Alliance dataset as readily usable by anyone as we can. The data is collected by thousands of people, in thousands of different environments, using many different methods. Such citizen science at scale is a huge achievement, but does come with a set of challenges, and one common piece of feedback we receive relates to the quality of certain data points. For example: the data on brands, which contains many typos and misspellings. With over 300,000 records containing a brand value, checking them all for correct and consistent spellings is no simple task. This year we have started initial work to improve the brand values in the released data, by tidying them up in the data from the largest data provider – The Restart Project. Here we made about 5000 corrections of 500 common misspellings of 50 common brands. These data quality tidyups are ongoing, and next year we aim to extend the brand corrections to the data of other data providers, as well as making recommendations to partners for reducing misspellings at the point that the data is captured.

For more information on how we produce the ORA dataset, see Producing the Open Repair Alliance datasets.

New this year: row by row data on non-EEE items

As well as new sources of data, this year we also have new types of data in the dataset.

Last year, we started experimental work on the inclusion of non-EEE items in the ORA datasets, providing top-level counts of categories and repair outcomes of non-EEE items. This year, we’ve gone a step further and are now providing a full download of these non-EEE items – 203,591 of them. Anyone can download the data on non-EEE repair attempts – you’ll find them in the download file.

There’s more work to come on non-EEE data – such as further investigation of categorisation, and the inclusion of barriers to repair. If you’re interested in non-EEE data and would like to shape how it evolves, get in touch.

A quick look at the data

There’s a lot of information in the Open Repair Alliance dataset, and it has been used in a variety of ways – you can view a highlight of the published reports and analyses on our Insights page, and you can explore the latest data in an interactive overview.

Some key takeaways from this year’s dataset:

• Items seen:
  • 619,875 items; from 44,252 events; by 1,728 groups; in 34 countries.
  • 416,284 EEE items (67%); 203,591 non-EEE items (33%).
  • Most seen:
    • Vacuum cleaners (34,840 seen), lamps (29,158), power tools (28,291) and coffee makers (24,651) are the most commonly seen EEE items.
    • Clothing/textiles (102,241), household items and tools (60,275), bicycles (18,162) and jewellery (12,662) are the most commonly seen non-EEE items.
• Repair success rates:
  • 399,776 items repaired; a 64% success rate.
  • 54% EEE items and 85% non-EEE items are fixed during an event.
  • Most success:
    • Sewing machines (70% fixed), lamps (70%), hair dryers (64%) and decorative or safety lights (63%) are the EEE items with the highest successful repair rate.
    • All non-EEE items have a high success rate (over 75%).
  • Least success:
    • Flat screens (31% fixed), DSLRs/video cameras (34%), digital compact cameras (38%) and printers/scanners (41%) are the items with the least successful repair rate during events.
• Barriers to repair:
  • Of those EEE items that are deemed to be end-of-life during an event, the most commonly recorded reason is: spare parts not available (12,681).

We’ll publish a detailed post soon going into more detail on the different ways in which you can use the data. Stay tuned!

Calls to action

There’s lots of ways you can get involved with the Open Repair Alliance data.

• Download the data and tell us what you find.
• If you’re a group, start logging the data from your events and share it digitally.
• Take our survey and give us your feedback.

And don’t forget – tomorrow (17th October 2026) is International Repair Day!

The post appeared first on Open Repair Alliance.