Top Banner
Aimée Vargas, Jyh-Ming Lien, Marco Morales and Nancy M. Amato Algorithms & Applications Group Parasol Lab, Dept. of Computer Science, Texas A&M University VIZMO++ A Visualization, Authoring, and Educational Tool for Motion Planning
14

Aimée Vargas, Jyh-Ming Lien, Marco Morales and Nancy M. Amato Algorithms & Applications Group Parasol Lab, Dept. of Computer Science, Texas A&M University.

Dec 14, 2015

Download

Documents

Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Aimée Vargas, Jyh-Ming Lien, Marco Morales and Nancy M. Amato Algorithms & Applications Group Parasol Lab, Dept. of Computer Science, Texas A&M University.

Aimée Vargas, Jyh-Ming Lien, Marco Morales and Nancy M. Amato

Algorithms & Applications Group

Parasol Lab, Dept. of Computer Science,

Texas A&M University

VIZMO++

A Visualization, Authoring, and Educational Tool for Motion

Planning

Page 2: Aimée Vargas, Jyh-Ming Lien, Marco Morales and Nancy M. Amato Algorithms & Applications Group Parasol Lab, Dept. of Computer Science, Texas A&M University.

• Motion Planning consists of finding feasible motions for

movable objects• Many applications: robotics, computer graphics, scientific

computing, …• Input

− Data describing the robot’s shape, articulations, constraints− Data describing the shape and configuration of obstacles − The robot’s start and goal configurations

• Planning Algorithm (e.g., sampling-based motion planning)− Samples configurations randomly in the Configuration Space (C-Space)− Connects each sample to one or multiple near-by samples

• Output− A graph describing the feasible motions− A sequence of configurations that take the robot from the start to the goal

Algorithmic Motion Planning

workspace

Page 3: Aimée Vargas, Jyh-Ming Lien, Marco Morales and Nancy M. Amato Algorithms & Applications Group Parasol Lab, Dept. of Computer Science, Texas A&M University.

Algorithmic Motion Planning: Input

4

MultiBody Active1FreeBody 0 small_robot.g 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000Connection0

MultiBody Pasive1FixedBody 0 tunnel.g 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000Connection0

MultiBody Pasive1FixedBody 0 block.g 16.000000 13.000000 0.000000 0.000000 0.000000 0.000000Connection0

MultiBody Pasive1FixedBody 0 block.g -16.000000 13.000000 0.000000 0.000000 0.000000 0.000000Connection0

1 36 76 228 1 76-12.500000 -2.500000 2.500000-9.500000 -2.500000 2.500000-9.500000 -2.500000 -2.500000-12.500000 -2.500000 -2.500000-12.500000 2.897685 2.500000-9.500000 2.897685 2.500000-12.500000 2.897685 -2.500000-9.500000 2.897685 -2.500000-4.500000 -2.500000 2.500000-4.500000 -2.500000 -2.500000-4.500000 12.500000 2.500000-9.500000 12.500000 2.500000-9.500000 12.500000 -2.500000-4.500000 12.500000 -2.500000-12.500000 -12.500000 12.50000012.500000 -12.500000 12.50000012.500000 12.500000 12.500000-12.500000 12.500000 12.50000012.500000 12.500000 -12.500000-12.500000 12.500000 -12.5000009.500000 12.500000 2.5000004.500000 12.500000 2.5000004.500000 12.500000 -2.5000009.500000 12.500000 -2.50000012.500000 -12.500000 -12.500000-12.500000 -12.500000 -12.50000012.500000 -2.500000 -2.50000012.500000 -2.500000 2.50000012.500000 2.897685 2.50000012.500000 2.897685 -2.5000009.500000 2.897685 2.5000009.500000 2.897685 -2.5000009.500000 -2.500000 2.5000009.500000 -2.500000 -2.5000004.500000 -2.500000 2.5000004.500000 -2.500000 -2.5000001 2 -4 2 3 -4 5 6 -1 6 2 -1 7 8 -5 …

Workspace file

1 8 12 36

1 12

0.95 0.95 0.95

0.95 0.95 -0.95

0.95 -0.95 0.95

0.95 -0.95 -0.95

-0.95 0.95 0.95

-0.95 0.95 -0.95

-0.95 -0.95 0.95

-0.95 -0.95 -0.95

1 5 -7

1 7 -3

1 2 -6

1 6 -5

1 3 -4

1 4 -2

5 6 -8

5 8 -7

2 4 -8

2 8 -6

3 7 -8

3 8 -4

robot obstacle

Geometry files

Page 4: Aimée Vargas, Jyh-Ming Lien, Marco Morales and Nancy M. Amato Algorithms & Applications Group Parasol Lab, Dept. of Computer Science, Texas A&M University.

Algorithmic Motion Planning: Output

Roadmap Version Number 061300#####PREAMBLESTART##### ../obprm -cd RAPID -f serial -bbox [0.0,4.0,0.0,4.0,-5.0,14.0] -bbox_scale 1 -lp straightline rotate_at_s -gNodes OBPRM nodes 400 collPair rT rT shells 1 freePair rT rT -cNodes closest 20 components 5 5 #####PREAMBLESTOP##########ENVFILESTART#####serial.env#####ENVFILESTOP##########LPSTART#####straightline rotate_at_s 0.5a_star_distance 6 9#####LPSTOP##########CDSTART#####3vclip RAPID PQP #####CDSTOP##########DMSTART#####5scaledEuclidean 0.9euclidean minkowski 3 3 0.3333manhattan com #####DMSTOP##########GRAPHSTART#####89 78 890 1.95659 2.51507 -4.62012 0.763875 0.92729 0.307945 0.984243 0.0868963 0.304818 1 -1 -1 2 4 1 122 3 126 1 2.87237 0.864196 -4.5569 0.675137 0.382797 0.266389 0.581803 0.195711 0.873085 1 -1 -1 1 8 1 133 2 2.9388 3.5578 -4.30382 0.115513 0.0309897 0.993182 0.688779 0.297365 0.0481244 1 -1 -1 1 12 1 223 3 3.00148 0.25514 -4.59405 0.0368278 0.900472 0.244192 0.909281 0.937958 0.433103 1 -1 -1 1 0 1 126 4 3.00187 0.345139 -4.55145 0.374121 0.167178 0.795772 0.606054 0.804405 0.825447 1 -1 -1 2 7 1 32 0 1 122 5 0.940853 0.879135 -4.65633 0.334515 0.538076 0.575211 0.163215 0.918796 0.106324 1 -1 -1 2 9 1 230 14 1 260 6 1.4835 3.37686 -4.45324 0.161255 0.324899 0.0697619 0.261815 0.247689 0.382626 1 -1 -1 1 24 1 227 7 2.82442 0.927611 -4.5542 0.226048 0.822072 0.33608 0.898007 0.667324 0.289052 1 -1 -1 3 4 1 32 9 1 219 24 1

251 8 0.34911 1.50508 -4.16536 0.994804 0.295583 0.435455 0.627437 0.361991 0.973598 1 -1 -1 4 1 1 133 13 1 202 24

1 202 23 1 237 9 2.20836 0.059729 -0.339691 0.215222 0.247623 0.330884 0.157593 0.796295 0.400327 2 -1 -1 2 7 1 219 5 1 230

......

#####GRAPHSTOP#####

VIZMO_PATH_FILE Path Version 20001125

1

1443

20.000000 0.000000 0.000000 0.000000 0.000000 0.000000

19.910263 0.013349 0.017175 0.994866 0.992220 0.996640

19.820526 0.026698 0.034349 0.989732 0.984439 0.993280

19.730789 0.040048 0.051524 0.984598 0.976659 0.989919

19.641052 0.053397 0.068699 0.979464 0.968878 0.986559

19.551315 0.066746 0.085873 0.974331 0.961098 0.983199

19.461578 0.080095 0.103048 0.969197 0.953318 0.979839

19.371841 0.093445 0.120222 0.964063 0.945537 0.976478

19.282104 0.106794 0.137397 0.958929 0.937757 0.973118

19.192367 0.120143 0.154572 0.953795 0.929976 0.969758

19.102630 0.133492 0.171746 0.948661 0.922196 0.966398

19.012893 0.146842 0.188921 0.943527 0.914416 0.963038

18.923157 0.160191 0.206096 0.938393 0.906635 0.959677

18.833420 0.173540 0.223270 0.933259 0.898855 0.956317

18.743683 0.186889 0.240445 0.928126 0.891075 0.952957

18.653946 0.200239 0.257619 0.922992 0.883294 0.949597

18.564209 0.213588 0.274794 0.917858 0.875514 0.946237

18.474472 0.226937 0.291969 0.912724 0.867733 0.942876

18.384735 0.240286 0.309143 0.907590 0.859953 0.939516

18.294998 0.253636 0.326318 0.902456 0.852173 0.936156

18.205261 0.266985 0.343493 0.897322 0.844392 0.932796

18.115524 0.280334 0.360667 0.892188 0.836612 0.929435

18.025787 0.293683 0.377842 0.887054 0.828831 0.926075

17.936050 0.307033 0.395017 0.881920 0.821051 0.922715

17.846313 0.320382 0.412191 0.876787 0.813271 0.919355

Path file Roadmap (graph) file

Page 5: Aimée Vargas, Jyh-Ming Lien, Marco Morales and Nancy M. Amato Algorithms & Applications Group Parasol Lab, Dept. of Computer Science, Texas A&M University.

Motivation Meaningful visualizations and interfaces to

– Illustrate Problems Solutions

– Help in the development and testing of motion planners

– Create and modify problem instances– Try queries

Teaching aid– Illustrate how planners model motions in complex

problems– Illustrate the many abstractions involved in motion

planning– Allow to deal with interesting environments that

otherwise may be impossible to access– Illustrate differences and similarities between motion

planners– Interactive learning

– Degrees of Freedom– Configuration– Configuration Space– N-dimensional Space– Graph, Tree– Connected Component– Medial Axis– Distance Metrics– …

Page 6: Aimée Vargas, Jyh-Ming Lien, Marco Morales and Nancy M. Amato Algorithms & Applications Group Parasol Lab, Dept. of Computer Science, Texas A&M University.

VIZMO++

The major components include: A user friendly interface

– C++, OpenGL and Qt 4 Visualization of the

workspace/scene Visualization of results

generated by motion

planners Editor for the

workspace/scene Editor for roadmaps

Page 7: Aimée Vargas, Jyh-Ming Lien, Marco Morales and Nancy M. Amato Algorithms & Applications Group Parasol Lab, Dept. of Computer Science, Texas A&M University.

VIZMO++

The major components include: A user friendly interface Visualization of the

workspace/scene– Wired/solid modes

– Change colors Visualization of results

generated by motion

planners Editor for the

workspace/scene Editor for roadmaps

Page 8: Aimée Vargas, Jyh-Ming Lien, Marco Morales and Nancy M. Amato Algorithms & Applications Group Parasol Lab, Dept. of Computer Science, Texas A&M University.

VIZMO++

The major components include: A user friendly interface Visualization of the

workspace/scene Visualization of results

generated by motion

planners– path sweep volume

– path animation Editor for the

workspace/scene Editor for roadmaps

QuickTime™ and a decompressor

are needed to see this picture.

Page 9: Aimée Vargas, Jyh-Ming Lien, Marco Morales and Nancy M. Amato Algorithms & Applications Group Parasol Lab, Dept. of Computer Science, Texas A&M University.

VIZMO++

The major components include: A user friendly interface Visualization of the

workspace/scene Visualization of results

generated by motion

planners– debugging

– compare different methods Editor for the

workspace/scene Editor for roadmaps

Page 10: Aimée Vargas, Jyh-Ming Lien, Marco Morales and Nancy M. Amato Algorithms & Applications Group Parasol Lab, Dept. of Computer Science, Texas A&M University.

VIZMO++

The major components include: A user friendly interface Visualization of the

workspace/scene Visualization of results

generated by motion

planners Editor for the

workspace/scene– select/translate/rotate/scale

– inset/delete objects

– check for collision Editor for roadmaps

Page 11: Aimée Vargas, Jyh-Ming Lien, Marco Morales and Nancy M. Amato Algorithms & Applications Group Parasol Lab, Dept. of Computer Science, Texas A&M University.

VIZMO++

The major components include: A user friendly interface Visualization of the

workspace/scene Visualization of results

generated by motion

planners Editor for the

workspace/scene Editor for roadmaps

– modify

– add

– delete

Page 12: Aimée Vargas, Jyh-Ming Lien, Marco Morales and Nancy M. Amato Algorithms & Applications Group Parasol Lab, Dept. of Computer Science, Texas A&M University.

Demo

Page 13: Aimée Vargas, Jyh-Ming Lien, Marco Morales and Nancy M. Amato Algorithms & Applications Group Parasol Lab, Dept. of Computer Science, Texas A&M University.

Conclusion

VIZMO++ is a 3D tool for visualizing and editing motion planning

– Environments and Problem Instances– Solutions

Future work includes– allow users to construct articulated robots– support multiple robots– support Constructive Solid Geometry (CSG)– Configuration-space visualization (2D-3D projections)– release the code

Page 14: Aimée Vargas, Jyh-Ming Lien, Marco Morales and Nancy M. Amato Algorithms & Applications Group Parasol Lab, Dept. of Computer Science, Texas A&M University.

Aimée Vargas, Jyh-Ming Lien, Marco Morales and Nancy M. Amato

Algorithms & Applications Group

Parasol Lab, Dept. of Computer Science,

Texas A&M University

VIZMO++http://parasol.tamu.edu/groups/amatogroup/research/vizmo++/