bug: lineal acceleration chart showing incorrectly

[Harry]

Not sure we are on the same page... LapTimer measures acceleration using the accelerometer. Bit due to the lean angle on a motorbike lat acceleration measured is zero. So the fallback is calculating it from GPS.

Harry

[i-Zapp]

Not sure we are on the same page... LapTimer measures acceleration using the accelerometer. Bit due to the lean angle on a motorbike lat acceleration measured is zero. So the fallback is calculating it from GPS.
Harry

on the exact same page. my hope was that this shortcoming would have been addressed for your motorbike customers so that it's no longer a 'fallback'. I say 'shortcoming' because in lieu of using the high-resolution accelerometer data available in the iPhone, the lateral acceleration is based on relatively low resolution (1 hz?) GPS data and subject to the inaccuracies of that position data.

no offense intended, and i genuinely appreciate the quality of your work and customer support. it just seems as though HLT is biased more towards cars than motorbikes or i must be underestimating the complexity of determining lean angle and capturing the vertical axis of acceleration.

[gplracerx]

i-Zapp,

It could be done using what's called sensor fusion (or inertial navigation) where you use all three accelerometer axes plus the three rotation rate axes to determine vehicle orientation, velocity and position. You then use GPS and the three axis magnetometer (if it's there) to correct for the drift from integrating the noise in the sensors. Drift is quite high since the sensors are noisy. You then can get Euler angles for the vehicle relative to the Earth reference plane. The roll angle would be the bike lean. For drifting, you could get the yaw angle relative to the velocity vector. It would probably also help for turns with a lot of banking like on the road course at Daytona or the Nordschleife for that matter. The DIY drone community has done this for flight control of model planes and helicopters.

Perhaps we'll see it some time in the future, version 19 maybe.

[i-Zapp]

i presume that the iPhone has a 3-axis accelerometer. for automotive applications, the vertical axis is not really used (assuming the phone is layed flat, otherwise you would calibrate that out). however, for motorbikes it IS one of the two primary axis: vertical for use in lateral acceleration along with the sine of the lean angle, and longitudinal for accel and braking. i think that part is easy. I suspect the lean angle is a bit trickier, which requires integrating the rolling movement of the phone/bike around its longitudinal axis. the devil's in the details, and i'm guessing Harry's not releasing the feature until its perfect.

[gplracerx]

The iPod Touch and other smart devices have three axis accelerometers and three axis MEMS gyros (really rotation rate sensors). The iPhones have built in GPS and a three axis magnetometer too. The primary use is to orient the display on the device screen. The magnetometer can be used in maps apps as well as a cute little app called Gyrocompass 3D ( http://vitotechnology.com/gyrocompass.html ). There are equivalent apps for Android devices as well.

If you're technically oriented, here's a reference on direction cosine matrices and sensor fusion as applied to flight control: http://gentlenav.googlecode.com/files/DCMDraft2.pdf
If you're really into the math, this one is much more detailed, ( there are lots of links to more related stuff too): http://www.starlino.com/dcm_tutorial.html