Differences

This shows you the differences between two versions of the page.

--- tutorials:intermediate:simple_mobile_manipulation_plan [2019/07/16 11:26] – [Some Useful Designators] ?grasp-pose -> ?grasp-pose-identifier amar
+++ tutorials:intermediate:simple_mobile_manipulation_plan [2019/08/02 14:31] – [Some Useful Designators] Added more information about the object designator amar
@@ Line 42: / Line 42: @@
       * <code lisp> (desig:a motion (type going) (target ?loc-desig)) </code> is a motion that moves the body of the robot and expects a key ''target'', whose value should be of type ''location-designator''.
       * <code lisp> (desig:a motion (type looking) (target ?loc-desig))</code> is a motion that moves the head of the robot to look at the specfied location. It expects a key ''target'' with value of type ''location'' designator.
-      * <code lisp> (desig:a motion (type detecting) (object ?obj-desig)) </code> is a a motion that perceives and detects the specified object. It expects a key ''object'' with a value of type ''object-designator''
+      * <code lisp> (desig:a motion (type detecting) (object ?obj-desig)) </code> is a a motion that perceives and detects the specified object. It expects a key ''object'' with a value of type ''object-designator''. An example is <code lisp>
+      (desig:a motion (type detecting) (desig:an object (type bowl))) </code>. This motion perceives and tries to find an object of type bowl in the field of view where the camera of the robot is pointed to. The resultant of this is also an object designator, but it has much more information. A sample is given below <code lisp>
+    #<A OBJECT
+    (TYPE BOWL)
+    (NAME BOWL-1)
+    (POSE ((:POSE
+            #<CL-TRANSFORMS-STAMPED:POSE-STAMPED
+   FRAME-ID: "base_footprint", STAMP: 1.564755796743144d9
+   #<3D-VECTOR (0.7672472770564228d0 -0.35741216776250767d0 0.888642116546631d0)>
+   #<QUATERNION (-0.005783746257358676d0 -0.0013544535828327665d0 -0.26923602469246516d0 0.9630558655492675d0)>>)
+           (:TRANSFORM
+            #<CL-TRANSFORMS-STAMPED:TRANSFORM-STAMPED
+   FRAME-ID: "base_footprint", CHILD-FRAME-ID: "bowl_1", STAMP: 1.564755796743144d9
+   #<3D-VECTOR (0.7672472770564228d0 -0.35741216776250767d0 0.888642116546631d0)>
+   #<QUATERNION (-0.005783746257358676d0 -0.0013544535828327665d0 -0.26923602469246516d0 0.9630558655492675d0)>>)
+           (:POSE-IN-MAP
+            #<CL-TRANSFORMS-STAMPED:POSE-STAMPED
+   FRAME-ID: "map", STAMP: 1.564755796743144d9
+   #<3D-VECTOR (1.5255411783854167d0 0.761638069152832d0 0.8886421203613282d0)>
+   #<QUATERNION (-0.005410938989371061d0 -0.0024511234369128942d0 -0.07722260057926178d0 0.9969961643218994d0)>>)
+           (:TRANSFORM-IN-MAP
+            #<CL-TRANSFORMS-STAMPED:TRANSFORM-STAMPED
+   FRAME-ID: "map", CHILD-FRAME-ID: "bowl_1", STAMP: 1.564755796743144d9
+   #<3D-VECTOR (1.5255411783854167d0 0.761638069152832d0 0.8886421203613282d0)>
+   #<QUATERNION (-0.005410938989371061d0 -0.0024511234369128942d0 -0.07722260057926178d0 0.9969961643218994d0)>>)))> </code> Note that this resulting object designator has the name of the object along with the coordinate of the object w.r.t the robot coordinate frame and other useful coordinate transformations which can later be used to calculate the pose of the end effector for grasping.
     * **Action Designators** - These describe the high-level actions which may consist of multiple calls to different low-level motion to carry out a small plan.
         * <code lisp> (desig:an action (type picking-up) (arm ?grasp-arm) (grasp ?grasp-pose-identifier) (object ?obj-desig)) </code> is an action that picks an object with the specified arm and grasp pose. It expects an ''object'' key with value of type ''object-designator'', an ''arm'' with the value of type keyword for the arm to choose, and a ''grasp'' key whose value specifies a keyword which is an identifier of the grasp pose to pick the object with.
@@ Line 360: / Line 384: @@
                                (object ?perceived-bottle)))
 </code>
 We see that the robot defaults the right arm when the object is in the center and will always try to grasp the left side of the bottle, even when the object is slightly favoring the left side under it (Refer the figure below).
+Note: ''We have been using left-side here a lot. In the context of (grasp left-side), it refers to the left-side of the bottle. But what defines the "left-side"?''
+''Before we understand it, we have to acknowledge that every object spawned in the bullet world has its own coordinate axes - even rotationally symmetric objects like the bottle here. And the definition for left-side is the side of the -ve Y-axis in the bottle coordinate frame. Similarly right-side is +ve Y-axis, front is denoted by +ve X-axis and the back by -ve X-axis. ''
+''With this the "left-side" grasp can change based on how the object is oriented in the world. But it remains the w.r.t the object coordinate frame.''
+''The file at <your workspace>/cram/cram_common/cram_object_knowledge/src/household.lisp holds some definitions for grasps supported by the household objects supporte in CRAM.''
 {{:tutorials:intermediate:btw-tut-grasp-config-fail.png?800|}}
@@ Line 492: / Line 523: @@
 You can now try spawning the bottle in different points on the table and observing how the robot resolves the arm and grasp for it. Subsequently, you will notice that our plans will work as long as the bottle is within reach of the robot. You could try implementing the adjustment of the robot's position to facilitate the fetch.
+Instead of a hardcoded pose for the base of the robot, you now would now need to obtain a list of poses where the robot will be able to grasp the bottle from. This is exactly where [[tutorials:intermediate:costmaps|location costmaps]] come into play. CRAM provides an easy way to handle these where the user does not have to bother with hardcoded poses. A way to obtain this would be through the following designator
+<code lisp>
+(a location
+   (reachable-for pr2)
+   (object (an object
+               (type bottle))))
+</code>
+Referencing this will give you locations where our robot PR2 can successfully reach an object of type bottle. Note that the object designator used here is pretty generic and will apply to all the bottles in the robot's belief state.
+To make it more specific we can provide the name and the location designator where the bottle is expected to be in.
+Eg:
+<code lisp>
+;; This is the location designator used to reference locations where the pr2 should be to reach an object
+;; named bottle-1 of type bottle which is situated on the counter-top named sink-area-surface
+(a location
+   (reachable-for pr2)
+   (object (an object
+               (type bottle)
+               (name bottle-1)
+               (location (a location
+                            (on (an object
+                                    (type counter-top)
+                                    (urdf-name sink-area-surface)
+                                    (part-of kitchen))))))))
+</code>
+You can see that the designator is not that intimidating as it looks, but provides a structured way to reference objects and locations. You can use these to write plans without having to worry about the actual coordinates.
 Since this is a simple tutorial in formulating and understanding mobile plans using CRAM, developing advanced plans and recovery behaviors is left up to you.
@@ Line 595: / Line 653: @@
---></html>
+-->
+</html>