|
|
|
Another 4 years of this
- Thread starter dragonrider
- Start date
I’ve spent my career in robotics – although not for much longer – and I believe that designing something for a zero-gravity environment would present many more problems than actual working conditions in zero gravity would solve.
I think there are a couple of key challenges.
Robotics, or in fact any electronics, designed to be deployed in space are subject to very constrained hardware requirements. Sensitive electronics must be protected from all types of radiation. This includes gamma rays, x-rays, ultraviolet radiation, visible light, infrared radiation and microwaves.
They must also be protected from extreme heat and cold.
Of course, an advantage of deploying a robotics in space is that do not have to withstand the forces and moments imposed by gravity (so in theory can be quite light), but this is a double-edged sword as the base is also relatively light. Movement in the arm will introduce reaction forces in the base, and so now the problem of managing dynamics, control and motion planning has grown exponentially as the base (Space Shuttle, IIS, etc.) must be considered.
Due to these additional dynamics, the end of the arm is much harder to control and can easily miss the target.
The problem is magnified when the arm grabs something that has a large mass, like a satellite. In zero-gravity the control program must take into account the movement of the arm and base in conjunction with the mass of the target and base. I would imagine its not easy.
I would think that this dependency between the base and the arm places many constraints on both and is something that is negligible in the robots we build as they are anchored to the earth.
There are some possible advantages, materials can be lighter, planetary gears can be smaller, but given the additional articulation I would imagine the bearings would have to be larger for a give arm diameter making less room for sensors, etc.
I personally don’t believe it would be easier but its an interesting engineering design question. Sorry, just thinking out loud.
Edit: While I'm on my Yeah Canada rant, one of the 2 most widely used systems for real-time process control is QNX from a company based in Ottawa. The other is VxWorks from California.
I started off working for a company called Kelk in Don Mills that do Sensors for Rolling Mills using QNX 2.15A. I'll never forget it
Kelk are still around despite my best efforts
Last edited:
dragonrider
Well-known member
dragonrider
Well-known member
|
|
|
|
|