Stryd has come along and set a new benchmark in data quality for fitness devices, in my opinion. At this point, I would not feel as confident in my training numbers if I were still relying on GPS for distance and pace, especially at a track, in the city, or on a trail.

Many misconceptions about GPS accuracy need to be addressed.  GPS has several real limitations, especially consumer GPS chipsets found in all common sport watches. They are great for showing your gross location on the earth, but not as good for precision tracking and measurement, unless you are using special equipment. You can buy/build a system that will give you 1-centimeter accuracy, however, you will need a stationary base unit and have a second unit to wear as you run. In no way is a singular GPS device triangulating your location from satellites over 12,000 miles away (GPS) more accurate than a calibrated foot pod, which reads your foot movement in real-time (assuming both devices are working correctly).

General info on GPS technology and what’s actually required for precision tracking with GPS (RTK, differential GPS, DIY RTK, etc) to be added.

Some general info on accelerometers and integral navigation to be added.

 

How can Stryd be so precise and accurate?

Possesses a 9-axis motion sensor (link to sensor whitesheet when available):

3-axis accelerometer – measures linear accelerations in x-,y-, and z-direction (likely at 30 hz or more – meaning 30 readings per second!)

3-axis gyroscope – measures rotation about x-, y-, and z-axis

3-axis magnetometer – monitors orientation with respect to Earth’s magnetic field to compensate for long-term drift errors

Altimeter/barometer – detects absolute pressure changes to estimate relative elevation differences (likely reads several times per second in 0.2 meter resolution)

 

Example:

Assume my average stride length is 1 meter and and my cadence is 180 steps per minute.

At 180 steps per minute, I step three times per second, giving me about 3 meters of distance covered per second (3 m/s = 6.7mph or 8:56 min/mile).

GPS

Let’s be generous and let consumer GPS accuracy be 3 meters (9.84 feet) per reading. Real accuracy is closer to 10 meters.

Average error per reading = +/- 3 meters (assume a circle surrounding your GPS watch that your location will fall inside)

Total distance ran per reading = 3 meters

Total error per reading = up to 3 meters or 300 cm per reading*

*This is a worst-case scenario with only a single reading. This error will be reduced as it is spread over thousands of data points, however your accuracy depends on many variables.

STRYD

My recorded accuracy for Stryd is about 0.45% for a measured distance.

Average error = 0.45% or 0.0045 per reading

Total distance ran per reading = 3 meters

Total error per reading = 0.0045 x 3 meters = about 0.0135 meters or 1.35 cm per reading

More examples to come…

 

My personal experience with GPS (in no way an appeal to authority on my opinion of Stryd vs. GPS):

Starting geocaching with handheld and bluetooth GPS recievers in 2006.

Owned several handheld GPS recievers, GPS-enabled cellphones, and other GPS-enabled devices.

Own three GPS-enabled fitness devices (Polar M400, Garmin Fenix , Recon Jet).

Had an undergraduate class in surveying, learned differential GPS and survey techniques.

Took a special California-specific survey exam for professional engineering license.

Perform GPS surveys with differential GPS once or twice a year at work.

Work with GPS and survey info daily as geotechnical engineer.