Hi Xert team,
I would like to suggest a feature that could help users who train both indoors and outdoors with different power sources, for example a smart trainer indoors and pedals or a crank-based power meter outdoors.
The problem is that many users have real discrepancies between power sources. A trainer and a pedal-based power meter may not read the same, and the difference is often not just a simple fixed offset. In some cases, the difference is relatively small at steady endurance power, but larger at higher intensities, sprints, or short anaerobic efforts, or the other way around. A single percentage correction can therefore be only a rough compromise.
My idea would be to create a calibration / learning mechanism inside Xert:
-
The user connects both power sources during an indoor ride, for example the trainer and the pedals.
-
Xert records power from both sources at the same time.
-
Xert compares the two data streams across different power ranges.
-
Based on that comparison, Xert learns a dynamic correction curve rather than a single fixed offset.
-
The user picks a reference power source. Future activities from the other power source could then be transformed automatically so they better match the user’s reference power source.
This could be supported by a ready-made Xert test template. For example, Xert could provide a structured workout with blocks at different intensities: endurance, tempo, threshold, VO2max, and maybe a few short high-power efforts. After completing the test with two power sources connected, Xert could build a personalized power-source correction profile.
Even better, the system could continue learning over time. If the user occasionally records rides with two power sources, Xert could refine the correction curve and make the transformation more accurate.
This seems to fit very well with Xert’s philosophy. Xert is probably the only platform that really tries to connect indoor and outdoor training in a unified way, especially with Magic Buckets and adaptive recommendations. But that also means that power-source consistency becomes very important. If indoor and outdoor power sources disagree, the fitness signature, breakthroughs, XSS distribution, and training recommendations can all be affected.
Right now, the practical options are not ideal:
-
apply a fixed offset, which may be inaccurate because the difference is often nonlinear;
-
use the same power source (e.g. pedals) on every bike and on the trainer, which is inconvenient;
-
manually adjust or ignore differences, which is not very Xert-like.
It would also be useful if Xert could recognize multiple outdoor power sources. Some users may have one power meter on a road bike, a different one on a gravel bike, another on an MTB, and then a different one on a trainer indoors. Ideally, Xert would detect the source device from the activity file and apply the appropriate dynamic correction profile for that specific source.
I think this could solve a real problem for users who train seriously across indoor and outdoor setups, and it would make Xert’s modeling even more robust.