Pass IMU accelerations from the simulator

[jonaraphael]

Right now we are only getting the imu orientation and angular rates. the function that lets us know if the robot might fall over needs the acceleration.

[iamgohan]

I'm sorry for going completely off topic here but this is the only place i can get in touch with the designers without involving kickstarter. With complete respect, i have to tell you that I believe you are approaching the design wrong, and wasting energy in the process. The "idea" you are looking for with this project going open source, is to simply toss the idea of "less weight is better" out of your mind. a counterweight on the legs inside the "shell" of the spider would allow you higher speeds and more maneuverability with less energy consumed.
It is simple, try to walk without slightly tilting your body from left to right and consciously force every movement of your legs. you can walk this way. you not only looks silly but you are unstable and it takes more energy. now walk normally and pay attention to the way your stride is powered..... during most of the stride you leg is limp and just simply swinging forward as a counterweight to your upper body. ( this is only slightly noticeable but it is there)

Imagine you have a robotic arm that weighs 3 tons, having it lift to shoulder height multiple times without a counter weight of some type would consume a great deal of power. if it was counterweighted properly a 5 year-old girl could move it. the same idea can be applied to the legs of you spider.
yes there are ways to move heavy objects like a counterweight relatively quickly
yes it would take tweaking to make sure the weight is in the proper locations as the leg extends ( the center of balance will change)
yes it is a completely changed design from what you are working on.
dealing with the problems caused by the momentum is something i do not want to even start thinking about.
but i beg of you, please divert from the status quo, stop doing what has already been done. you are funded by people that are looking for something new. people that have hope in the hearts for actual answers coming from science. people that look at the mess that is this world and hope science will come to the rescue. thats all i can say.

[becomesaflame]

It doesn't give us linear rates/ acceleration?

[jonaraphael]

Hello Gohan,
Thank you for your constructive input! I'm afraid you have only reached the code developers by posting to this site, so I recommend that you send any other ideas and questions to info@projecthexapod.com.

However, I am happy to respond to this one comment here. The problem with the approach you suggest in our set up is that we are on the very delicate edge of the actuator ability and structural ability of the materials. After a certain size is surpassed, it becomes a runaway problem, where adding more weight requires the legs to be stronger and beefier, which then adds to the weight. We are at the cusp of this positive feedback limit, and so adding any extra weight to the robot could be catastrophic, especially if it were on par with the mass of the legs.

It turns out your idea is excellent when applied to structures that are bolted to a fixed object, like a crane (counterweight!), or a robotic arm. Unfortunately, it is simply untenable at this scale on a moving robot, that has to support its own weight.

I hope this addresses your concern. We appreciate that you are also trying to change paradigms in the robotics community! Check out other design decisions we have made (and the reasons behind them) at http://projecthexapod.com/blog/presentations/ . And send any further response or suggestions you might have to the info@projecthexapod.com address.

Thanks again,
Jona