3
$\begingroup$

While doing a literature review of mobile robots in general and mobile hexapods in particular I came across a control system defined as "Task level open loop" and "Joint level closed loop" system.

The present prototype robot has no external sensors by which its body state may be estimated. Thus, in our simulations and experiments, we have used joint space closed loop (“proprioceptive”) but task space open loop control strategies.

The relevant paper is A simple and highly mobile hexapod

What is the meaning of the terms "joint-level" and "task-level" in the context of the Rhex hexapod?

$\endgroup$
  • $\begingroup$ I would appreciate a detailed question with specific problems. $\endgroup$ – Josh Vander Hook Apr 30 '13 at 13:19
  • $\begingroup$ I hope this edit is sufficient. I am pretty confused as to what sort of details I should provide. Especially since this was part of a literature review. $\endgroup$ – Naresh Apr 30 '13 at 13:30
  • 2
    $\begingroup$ First: Put a question in the body. Then: Summarize the question in the title. Here, I'll do it for you. $\endgroup$ – Josh Vander Hook Apr 30 '13 at 14:05
  • $\begingroup$ Thanks Josh and Shahbaz, your edits have improved this question substantially. Naresh, you may want to read How to Ask for tips on how to help make your questions better in the future. $\endgroup$ – Mark Booth Apr 30 '13 at 17:13
  • $\begingroup$ +1, those edits were a big help. $\endgroup$ – Ian Apr 30 '13 at 17:58
7
$\begingroup$

These are just terms to describe the "layers" of control on the robot. The "joint level" means the position of each actuator (leg), and the "task level" means the current goal of the robot (like go forward, go east, go to location X, etc).

This paragraph is about sensing. There are (apparently) position sensors in all of the leg joints, so the robot is capable of closed loop control at that level -- no user intervention needed. ("Proprioceptive" is a $5 word for "I know where my legs are because I can sense them".)

However, having no external sensors means that the robot can't tell where it is in relation to the world and as a result it can't determine what action it should take to accomplish its goals. So, it must run as open loop. Presumably this means that they are either giving it a pre-scripted set of actions, or directing it via remote control.

So, the task-level control handles the "go forward" concept, but the joint-level control handles the uneven terrain underfoot.

| improve this answer | |
$\endgroup$
  • $\begingroup$ servos typically go to a position that the electronics specify. so likely there are no sensors, just faith that the actuators are moving to the position they are told to. of course mechanical resistance could work against them. the 'proprioception' is implied by where the robot 'expects' the servos to be. $\endgroup$ – Octopus May 6 '13 at 22:11
  • $\begingroup$ After the passage quoted in the question, they go on to say that "All controllers generate periodic desired trajectories for each hip joint, which are then enforced by six local PD controllers (one for each individual hip actuator)." PD control requires sensing. $\endgroup$ – Ian May 7 '13 at 18:28

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.