Roller coaster forces in Ferrari Land + Port Aventura

[0Kajuna0]

Hi, new member here . Thanks for all the info you guys share.

Not exactly a trip report, so if this is not the right place to post this, please let me know.

I went to Ferrari Land and Port Aventura, rode all the coasters and recorded them on a GoPro, then took out the gyroscope and accelerometer data and this is the result. Hope you find it interesting!

 

[Hixee]

A nice study, and reasonably well presented too.

How much filtering did you have to apply to smooth out the data? My experience of using accelerometers in real world applications was that the noise from vibration made the measurements hard to analyse on the first pass.

 

[0Kajuna0]

So the accelerometer updates 200 times per second, I averaged that to 25 per second to match the video framerate, then averaged 5 frames or so to have reasonably smooth bars and values. When I show the Vibrations metric, the onscreen dot movement is averaged to the framerate and the calculations in Hz are based on the raw data without smoothing. The same applies to the gyroscope, but it comes in 400Hz or so instead of 200. The camera is mouth-mounted, so I believe it portrays well the forces you (your head) would experience.

 

[Hixee]

Sounds like a sensible method to me. My usage of this sort of thing required real time (as-near-as-instantaneous-as-possible) filtering of the acceleration data, so it's something I've always been particularly aware of!

A nice study, would be interesting to see how this sort of thing compares to the design data - although who knows how you'd ever get hold of that!

EDIT: Did a bit of stalking of your YouTube channel, and this appears to be your first coaster project. Looks to me more like a chance for you to try out this type of analysis from the GoPro. What inspired you to pick Port Aventura (other than the obvious advantages of using coasters for this kind of thing)? Did the park help you at all?

 

[0Kajuna0]

I suspect someone is working on live usage solutions of these GoPro data too, not sure if that might be interesting for you.

I wanted to try the GoPro metadata capabilities in coasters, as you say for their obvious advantages. One that might not be self evident are loops. The gyroscope data allows me to keep the horizon flat (with some manual tweaks as there is no absolute orientation data) and that effect is much cooler in a coaster than, say, on a motorbike (it's common in MotoGP). PortAventura is the nearest big park to Barcelona, where I live, and all that was a good enough excuse to go there and try Red Force (which also added some big numbers to the metrics). The park did not actively help, other than not preventing this from happening. Happy to shoot in more parks if they want something similar

At some point I might get my hands on a Hero5 Black (this was a Session), which has a GPS sensor, and add speed, distance and 3D tracks o this kind of video. It's supposed to be very precise (18Hz), so the resulting shapes could be pretty.

 

[Gazza]

I love how you've finally given enthusiasts a way to quantify roughness.

But, awesome videography. I don't think I've seen a video done this well in a long time.

 

[HeartlineCoaster]

Very interesting. Hope you can make it to many more places for this.

I love how you've finally given enthusiasts a way to quantify roughness.

Decibels from pain would be a suitable alternative.

 

[Hixee]

I suspect someone is working on live usage solutions of these GoPro data too, not sure if that might be interesting for you.

I used this kind of thing for my masters project many years ago now, so unfortunately (or fortunately, depending on your point of view) it's not something I use much anymore.

At some point I might get my hands on a Hero5 Black (this was a Session), which has a GPS sensor, and add speed, distance and 3D tracks o this kind of video. It's supposed to be very precise (18Hz), so the resulting shapes could be pretty.

Yeah, it would be interesting to see how this would work. What did you do your visualisation in?

Thanks for sharing this, really cool to see.

 

[Intricks]

This is actually really cool and interesting to see all the different type of forces we feel whioe roding a ride.

Also, too mention this again as posted by Gazza - you have given a visual representation of roughness based off data and not just experience (although we all experience pain differently).

Sent from my SM-G920P using Tapatalk

 

[Will]

I enjoyed that - very different and quite interesting.
I also like that roughness is now measured in Hurts (Ba-dum-tsch)

 

[0Kajuna0]

Thanks everyone for the kind words!

Yeah, it would be interesting to see how this would work. What did you do your visualisation in?

I was thinking of using Processing but ended up making my calculations there and producing the visualizations in After Effects, with most of the video filters and so on.
If anyone else has a Hero5 and would like to experiment with this stuff, I documented some of it here (edited, the link was wrong). The easiest way is to download the ZIP I put together and drop your video files on the batch script. Then you can visualise the data in Excel or anything similar, which can already produce interesting graphs. For example, this is Shambhala's vertical G-Force (without smoothing)