Marco Pavone
Stanford University
Notes by Paul Fischer
ReachBot: A Small Robot for Large Mobile Manipulation Tasks in Martian Cave Environments
ASL BDML earth and planetary surfaces group
Problem increasing reach and strength usually scales poorly with mass and complexity
Small robots are limited to small reachable wrkspace and wrench capability
Increasing reach and strength usually scales poorly with mass and complexity
Small robots are limited to small reachable workspaces
Enabling technology including extendable booms as controllable prismatic joints
Solar sails, panels, or camera booths
These booms have strong concise time
These can achieve extreme …
Feasibility challenges
Specific context of a martian cave exploration, focusing on noachian era targets
Maximize reachable and wrench workspaces
Positions shoulders and maintenance…
Creepers must be high weight, and must adapt and be both small and large
Control in motion changing policies
We aim to redeem robust control and motion planning
Finally natural mission for exploration and sampling of martian caves
Reliable lightweight surface grasping solutions
Inspiration from the fee of dexterous directionality
Instead of having fingers, we would use booms that provide pull
First hardware prototype provides unlimited banding and booms for the space structures
Retracting anchor arms allow the booms to provide movement
Early finding from the prototype using force control to prevent buckling and minimize oscillations
System level design considerations , degree of freedom for the consideration of the joints
How reachbot would explore the geometry of the given cave’s geometry
Do feasible trajectories exist then iterate on the process
This process will eventually lead to the second generation
Q: payload
A: the robot will have a number of sensors including cameras and microscopes, this is quite scalable that n engineer is looking at the tradeoffs… the instruments that have been discussed remain a question of tradeoff and imagine cameras and hypo spectral cameras in the near infrared and the thermal infrared miniaturize it enough to fit on the robot
Q: compatible with robotic armor in caves
A: asteroids and comets and in the context of assistance robots for a number of tasks, we are focusing on the caves
Q: grippers for all surfaces or just for rocks?
A: made for rock surfaces and knew that other robotic … had consequences no confinement to working in rocky terrain, in adhesive grippers with smooth materials, terrain such as caves where the competitive advantages such as this can shine, the grippers should be light and take load comparable to what the booms can support, up to tens of kg at most. At the strength of the rock strippers as well…
Q: rock debris at the spooling of the arm?
A: definitely a big challenge for most concerned explorers, same challenge of dealing with dust and the focus of the investigation it is not, but in the future
Q: booms more efficient from the ground?
A: motivation for the cavelike surfaces is to get up to the ceiling and higher surfaces with less debris
Q: structure overall a tethered unit or stand alone?
A: no tether, carry the subsystems making heavier and in terms of reach and range, currently we are exploring the design phase it could be tethered or untethered
Q: how to determine the appropriate anchor points?
A: the proposed strategy is a semi autonomous one, id candidate grip sites and the idea is to have sensor s at the gripper and see if this is a good feature for grasping. On the control planning and control side, a combination of careful selection of anchors and redundancy like climbers
It is the nature of this sort of climbing that a sort of grip will fail. Tether can be used a s a structural member and reaches the bottom with the grippers and booms and up towards the walls, an intention to keep the boom as much as possible
Q: how to handle unstable rock grips and pile, is there an autonomous capability?
A: in the end, you don’t really really know just like a rock climber, if it turns out to be loose and it comes off, you would be able to move fairly quickly without a new grip and slowing down the whole process
No comments:
Post a Comment