Sensor Placement Guide
Where to put each sensor on the body, and the reasoning behind it. The first section is a quick reference for the standard setups. The sections after it cover every supported location and the principles that make placement reliable, for users building their own multi-sensor configurations.
Quick reference: standard setups
Walking and Timed Up and Go
- One sensor on the top of each foot (see “Feet” below).
- Optional: one sensor on the lower back for trunk motion.
Balance and sway
- One sensor on the lower back, on the midline (see “Lower back” below).
Full lower body
- One sensor on the top of each foot, each shin, and each thigh.
- One sensor on the lower back or pelvis.
Every trial should begin with a few seconds of quiet standing (or sitting) so the sensors can establish which way is down. Keep still until the recording is under way.
Where each sensor goes
Sensors attach with the provided straps or belt. The mounting position sets which body segment a sensor measures.
Feet. Place the sensor on the top of the foot (the instep), over the midfoot above the toes, secured across the laces or with the foot strap. This dorsal midfoot position is the standard location used across clinical and research gait systems, so data is comparable with those systems.
Shins (lower leg). Place the sensor on the flat, bony front of the shin, on the rigid part of the bone rather than the soft calf muscle, so it does not wobble.
Thighs. Place the sensor on the front or outer thigh, snug enough that it moves with the leg and not with loose tissue.
Lower back. Place the sensor on the midline of the lower back, over the base of the spine, held by an elastic belt with the buckle at the front. Keeping it on the centre line, not off to one side, is important: an off-centre sensor mixes turning and side-to-side motion into the signals and reduces accuracy.
Pelvis. As an alternative to the lower back, a sensor can sit slightly lower over the pelvis or sacrum, on the same midline, with the same belt.
Chest. Place the sensor on the breastbone, centred on the flat part of the upper chest below the collarbones.
Upper arms. Place the sensor on the outer upper arm, midway between shoulder and elbow.
Forearms. Place the sensor on the back of the forearm, or on the wrist like a watch for a wrist setup.
Head. For head-mounted use, secure the sensor centred and level with a headband.
Principles
Orientation matters more than exact position. For measuring how a body segment moves, how the sensor is angled on the segment matters far more than the exact spot it sits. A sensor a couple of centimetres higher or lower on the same segment gives nearly the same result; a sensor twisted or tilted does not. This is why the brief quiet hold at the start of each trial matters: it lets the system learn the sensor’s angle and cancel out mounting tilt. It is also why placement does not need to be measured to the millimetre.
Keep trunk sensors on the midline. For the lower back, pelvis, and chest, the centre line keeps the sensor near the body’s line of rotation, which keeps turning and side-to-side signals clean.
Same placement for repeat visits. For tracking a person over time, put the sensors in the same spots each visit. Consistency between sessions matters more than any single exact location.
Notes on standards
Joint angles are reported using the segment and joint coordinate conventions recommended by the International Society of Biomechanics (Wu et al., 2002, for the ankle, hip, and spine; Wu et al., 2005, for the shoulder, elbow, wrist, and hand). There is no formal standard that dictates where wearable sensors are physically placed on the body; the positions above follow established clinical and research practice, and match the dorsal-foot and lower-back setups used by common commercial motion systems. The relationship between each sensor and its body segment is established by the quiet-standing calibration at the start of a trial, not by exact placement.