At House of Runners, we occasionally employ sensors strapped to the shoes of runners to understand the variables associated with a runner’s gait. This follows from my vision of using sensors and science to improve running, because the House is first and foremost a performance running club.
The specific device we use is called Runscribe, a 3D inertial measurement unit that Tim Clarke lent me at a subsidized cost when the early prototypes had come out in 2017. I was one of the lucky few who extensively tested them for him, relaying back comments and feedback to improve the metrics and device function. We’ve come a long way since the early designs.
Runscribe uses Invensense’s 9-Axis MPU-9255 chipset and a Nordic nRF52832 chipset to transmit low power Bluetooth signals. Specification is nearly 500Hz sampling frequency for raw Sensor data and 200Hz for Quaternion, Quaternion + RawData. Suffice to say, it’s a convenient device for field research applications.
Today’s featured runner is Rawad, who volunteered to have his right foot instrumented during our Monday training run in April. Rawad is a recent comer to running and has been training consistently in his bid to get fit for high altitude mountain climbing this summer. Rawad ran his first 3000m track race just last month.
To start off, his Monday training modality was 6 x 500m with a time target of roughly 2:30 on each interval, indicating that he would have run at around 5:00 min/km give or take. Below, tabulated variables are provided separating them in terms of efficiency metrics, shock metrics and motion metrics. Note that the “population range” column provides the range of values seen in similar runners of similar gender running at approximately the same paces. The reason to discriminate in such a way is that all gait variables are strongly associated with a given running pace.
Unfortunately, we didn’t have the time to strap a device onto his left shoe so all the data below is specific to one foot. In each of the 6 intervals, we understand the following metrics :
|Efficiency Metrics||Value||Population Range|
|Step rate||170 per minute||170 – 180 per minute|
|Stride length||2.6 meters||2.2 – 2.4 meters|
|Contact time||290 – 310 milliseconds||275 – 325 milliseconds|
|Flight ratio||11 – 14%||–|
|Shock Metrics||Value||Population Range|
|Shock||11 – 14 G’s||–|
|Impact G’s||9 – 10 G’s||Average value < 12 G’s |
Range = 8 – 14 G’s
|Braking G’s||8 – 10 G’s||Average value = 10-10.5 G’s |
Range = 7.5 – 12.5 G’s
|Motion Metrics||Value||Population Range|
|Footstrike Type||Heel Srike||Average = Mid-foot strike|
|Pronation Excursion||-2 to -4 degrees||Average = -3 degrees|
|Max Pronation Velocity||400 to 450 degrees/second|| Average = 450 degrees/second|
The Shoeprint is a cool visualization of the motion the foot goes through from footstrike to toeoff. Every single step of the run is made visible. In the image above, I’ve isolated the Shock variable (in terms of G’s) for each footstrike. We note that heel running is predominantly visible and most of the shock is concentrated around the heel region.
Secondly, Rawad’s values are mostly concentrated on center line and slightly to the right of shoe centerline. Given that this is the right foot, it is possible that what we observe is just how the right foot rolls naturally to the right across the entire period of footstrike. This might also shoe that high pronation towards midline of the body is absent, which maybe a good thing.
The visualization might show the natural variance in his running style and nothing more. Still, very interesting.
Remarks and Cues for Improvement
- Stride rate : Rawad tends to take more contact time on each landing and a higher stride length between consecutive landings on the same foot. These variables are probably associated with his specific running pace, but there is some room to decrease contact time to sub-290 milliseconds by increasing stride rate into the 175 – 180 range.
- Contact time : Experienced long distance runners show a contact time betwen 190-250 milliseconds, however this variable is strongly associated with running pace. There does seem to be some room for improvement by effecting the stride rate and footstrike type.
- Footstrike : A shorter ground contact time is associated with a mid-foot strike, therefore trying a new variation of landing and toe-off is suited for experimentation. Heel strikers tend to have more knee and hip injuries as per research studies, although the research also indicates that on average, injury rates show little difference between heel and forefoot strikes. Do what you’re comfortable doing.
- Shock and Impact G’s : Shock is the amount of impact absorbed during footstrike. Rawad maybe able to reduce the impact G’s to sub-10G levels by making a change to his foot strike and stride rate, although all the variations in values I see are stable. Do note that a feeling of ‘heavier foot landings’ is possible when the shoe is worn out from excess mileage. Look for signs that a shoe change in imminent. In general, a running shoe lasts anywhere from 400-600 kms before it’s midsole structure is compromised.
- Pronation Excursion and Velocity : This is the degree to which the foot rolls as it lands and takes off and is an indicator for good shock absorption. Excessive pronation angles may lead to overuse injuries while running, therefore, it is prudent to continue to monitor this with change in fitness and running speeds. When continuing to monitor, if pronation angles are excessive (over-pronation) and the foot gives signs of excessive motion, specfic types of shoes and custom orthotics maybe pursued. However, it is too premature to state now whether Rawad’s values are in anyway above “normal” ranges.
- Balance and symmetry : No judgement can be given on symmetry between the two feet as the device was only worn on one shoe. We hope in future to instrument both feet to understand bilateral symmetry.
Thanks for reading. Please direct any questions on this post to me.