- 1. A major leap forward in automatic and manual selections
- Before
- Now
- Optimization through change tracking
- 2. Optimized calculations and pre-fills
- Exceptions for calculations and pre-fills
- 3. A closer look at offline mode
- 4. In summary
In an industrial environment, every second counts. We know how essential the smoothness and speed of our tools are for your daily operations. This is why we have deeply worked on optimizing the performance of our platform, both in online mode and offline mode.
These improvements already translate into significant time savings for our users — and even more for your teams in the field.
1. A major leap forward in automatic and manual selections
The biggest performance boost comes from the automatic and manual selection blocks.
Before
- To display 10 assets with 10 characteristics each, the system executed 10 API requests to load the assets, then 10 API requests per asset for each characteristic → a total of 110 API requests.
- With each request taking around 0.2 seconds, this meant 22 seconds of loading time (in online mode).
Now
- To display 10 assets with 10 characteristics each, only the 10 main requests are needed to load the information → a total of 10 API requests.
- With each request taking around 0.2 seconds, the result is just 2 seconds of loading time.
➡️ Result: almost instant display, with data already known by the server. Recalculations or updates are only performed when necessary (for example, if a value is modified).
Optimization through change tracking
From a technical perspective, here’s how these optimizations work:
- A field called “is_modified” has been added to each piece of data processed in our tools. This field can be:
- true, meaning the value was entered or modified by a user,
- false, meaning the value has not been modified by a user or is automatically calculated.
- For any data with “is_modified” = false, the system will not recalculate or reload this data from the database.
Concretely, since the launch of Ermeo:
- 174 million pieces of data from interventions have never been modified,
- 67 million pieces of data were actually entered by operators.
2. Optimized calculations and pre-fills
- Calculated values (example: C = A + B) are now only generated when they are actually needed, instead of systematically.
- All default values and pre-filled fields (texts, numbers, dates, multiple choices, etc.) also benefit from the “is_modified” logic, meaning they are only computed and loaded when necessary.
➡️ Example: if a calculation C = A + B has never been modified, the system will not recreate it but simply display it.
Exceptions for calculations and pre-fills
Some data continue to be handled as before, such as:
- Tasks linked to intervention attributes when real-time reporting is enabled,
- User attributes (needed for intervention transfers or email sending, for example),
- Current dates (which must be recalculated on the day of the intervention).
3. A closer look at offline mode
Offline mode also benefits from these improvements, though the gains are less striking than in online mode.
- Older devices see a clear improvement because fewer insertions are made into the local database,
- Pre-filled tasks — previously the main source of frustration in offline mode — are now much faster thanks to optimized pre-fills.
4. In summary
These optimizations deliver:
- A considerable speed boost for selections and pre-fills,
- A smooth experience, even on older devices,
- A massive reduction in loading times, equivalent to several hundred days saved for all our clients.
These improvements reflect our commitment to making our tools ever more efficient, reliable, and aligned with your industrial challenges.