The Floor-Level Camera: What Can You Actually Assess Remotely?
Priya, 34, hired you for 12 weeks of virtual coaching to prepare for her first 10K. She trains in a 3-by-4-meter apartment living room. At your first video session her laptop sits on the floor, so the frame shows only her shins and feet; the ceiling light is behind her and the connection drops to roughly 240p for a third of the call. Her wrist wearable reported a resting heart rate of 52 bpm yesterday, but it jumped between 48 and 96 during the same five-minute seated period. She sits nine hours a day, describes "tight hamstrings and a cranky low back," and owns one light band and a chair. She reports no pain at rest, no diagnosed medical condition, and has not exercised in 14 months. You have 45 minutes total and a second call booked in six days.
1. Given the camera position, lighting, and bandwidth, which components of a standard static posture and overhead squat assessment can you validly collect in this session, and which must be deferred? How do you restructure the 45 minutes?
Read the worked answer
What the frame supports: a floor-level, front-on shot at 240p shows foot and ankle position, base of support, heel lift, and frontal-plane knee tracking. Those observations are usable, but only as screening flags, not findings.
What it cannot support: head and neck position, shoulder elevation, thoracic rounding, lumbar and pelvic position, depth relative to femur, trunk lean, and arm position overhead - all are outside the frame. Low resolution also makes small-joint positions unreliable even when visible. Concluding "her arms fall forward because of tight lats" from this view would be guessing.
Restructuring the 45 minutes:
- Spend the first 10–12 minutes fixing the measurement environment, which is itself the intervention: laptop on a counter or stacked books at hip height, three meters back, window or lamp in front of her rather than behind, and a bandwidth test. Ask her to wear shorts and flat shoes and clear the floor.
- Use the middle block for subjective and health-history data, a PAR-Q review, and a talk-test cardio baseline - none of which depend on video quality.
- Capture only what the current angle can honestly show: a 30-second chair stand, single-leg stance, and five bodyweight squats filmed from the front. Note heel lift and knee tracking.
- Assign the deferred items as asynchronous homework: a front and side video of a five-rep overhead squat, filmed with the phone propped at hip height.
Document the deferral explicitly, with the reason, so the gap is a planned limitation rather than a missed assessment.
2. How should you treat the wearable heart-rate data in your baseline and programming decisions?
Read the worked answer
A wrist optical sensor reading 48–96 bpm while the client is seated and still is not a measurement - it is artifact. Optical sensors are highly sensitive to poor skin contact, arm position, motion, and ambient light, and a 48-bpm swing within five minutes exceeds anything physiological in a resting subject. Treat the whole trace as unusable rather than averaging it into a "resting HR of 52."
That matters because a single resting value anchors every downstream zone. If you plug 52 into a percentage-of-max scheme, you inherit an error you can never detect later. Two further cautions: (a) 52 bpm would be unremarkable in a trained endurance athlete but is worth a second look in someone who has not exercised in 14 months, so it is a prompt to re-measure, not a finding; and (b) age-predicted maximum heart rate (220 − age = 186 here) has a standard deviation of roughly 10–12 bpm, so the zone math is already ±20 bpm wide before the wearable error is added.
Practical plan: use the talk test and RPE as the primary intensity regulators for the first four weeks - they are valid, free, and unaffected by bandwidth. Ask her to record resting HR on three consecutive mornings, seated for five minutes, arm still, same wrist, before getting out of bed, and to send the screenshots. If the numbers stabilize, use them as a trend line, not as a zone calculator. If she wants real HR-based work later, recommend a chest strap for a submaximal step protocol.
3. What do you document from this session, and what specific re-assessment plan do you set before the six-day follow-up?
Read the worked answer
Document in a structured note: date, duration, technology conditions (device, bandwidth, camera height, lighting), what was observed and by what method, what was deferred and why, all subjective reports verbatim ("tight hamstrings," "cranky low back"), the health-history screen, and the plan. Record the wearable data as "unreliable - artifact suspected," not as a value. Note the consent conversation if any part of the session was recorded.
Re-assessment plan for day six:
- Setup checklist sent in advance: phone propped at hip height, three meters away, light source in front, flat shoes, shorts, clear floor, wired connection or hotspot if available.
- Front and side static posture, standing relaxed.
- Overhead squat, five reps, filmed front and side, with her counting reps aloud so the audio lag does not corrupt the timeline.
- Single-leg stance, 30-second chair stand, band pull-apart, and a push-up to a comfortable stop - all measurable and all visible at hip height.
- Talk test during a six-minute walk to establish a conversational threshold.
Set the expectation explicitly: the first session produced a partial baseline, and the second completes it. Tell her the reason - a bad camera angle produces bad data, and bad data produces a program that fixes problems she does not have. That framing protects adherence better than apologizing for the delay.