Movement & Vision Analysis
Movement & Vision Analysis reads your sailing video and turns what your body is doing into numbers. It looks at two things: how you move through a tack, and how active you are — and where you are looking — while sailing downwind. It runs automatically; there is no separate step to trigger and nothing to configure.
What it measures
Movement through a tack
This runs after you upload a session that has video attached. For every tack the analysis already detected from your GPS track, Movement & Vision Analysis pulls a roughly ten-second clip centred on the tack and estimates your body pose in each frame — a full skeleton of shoulders, hips, arms and legs.
From that it produces, for each tack:
- An annotated clip. The ten-second tack clip with the detected skeleton drawn on top, so you can see frame by frame how you came across the boat.
- Body-movement speed. The peak sideways speed of your body's centre of mass as you cross the boat, measured as a fraction of the frame width per second so it doesn't matter how far the camera was zoomed in. A higher number means a faster, more athletic crossing; a lower number a slower, more gradual one.
- Shoulder-crossing trace. Your shoulder position mapped from one side of the boat (−1) to the other (+1), plotted against time through the turn. Port-to-starboard and starboard-to-port tacks are flipped to read the same way and lined up at the moment of the tack, so you can lay every tack in the session over each other and see which ones you crossed early, late or in a rush.
You'll find all of this in the analysis view under Maneuvers: the clip and per-tack numbers sit in the tack list, and there are two comparison charts — body-movement speed against the speed you lost in the tack, and the shoulder trace of every tack on one set of axes.
Activity and vision downwind
This one runs live, on your phone, during a Training Mode session — it uses the front camera pointed at you while you sail. Roughly once a second the app estimates your body pose and the direction your head is facing, on device, alongside the usual GPS and sensor recording. It only does this on downwind legs.
Two figures come out of it, both measured as rates of change between frames:
- Body activity. How much your body is moving between one moment and the next — a proxy for hiking, pumping, moving your weight around and generally working the boat down the run.
- Vision. How much you are turning your head — a proxy for looking around to read the pressure and the puffs on the water rather than staring straight ahead. If your face isn't visible to the camera for a stretch, this is simply left blank for that stretch rather than counted as zero.
Training Mode uses both in its spoken commentary. At the end of each downwind leg — and in shorter updates roughly every 30 seconds during a long one — it compares the leg you just sailed against your own downwind average so far this session, for example telling you that you were looking around more than usual, or working the boat less than usual, on that run. It's descriptive, not instructional: neither more nor less movement or scanning is treated as "better", it just reports what changed. The figures are saved with the session alongside the commentary transcript.
Why single-handed only
Both pipelines analyse exactly one person. The tack analysis picks the largest person it detects in the frame; the downwind analysis tracks a single pose from the front camera. With two or more people on board, the metrics either jump between crew or lock onto whoever is most visible, and the numbers stop meaning anything. Until crew-aware tracking is added, Movement & Vision Analysis is only offered for single-handed dinghies, where there is one sailor to follow.
If you sail a keelboat or with crew and would be interested in helping us develop multi-person tracking, we'd like to hear from you — get in touch through our contact form with what you sail and what you'd want to see.
It also depends on you actually being in shot. For tack analysis that means the session video needs to show you through the tack; for downwind activity and vision it means the phone's front camera needs a reasonably clear view of you while recording a Training Mode session.
What you need
- Movement through a tack — a Pro account, and an uploaded session with a video that shows you on the boat.
- Activity and vision downwind — a Training Mode session recorded with the app, with the front camera able to see you. Processing happens on the phone; nothing extra is uploaded for it.