Robotics Research: Spatial Mechanism Design and Analysis and Autonomous Vehicle Development

1
Robotics ResearchSpatial Mechanism Design and
Analysisand Autonomous Vehicle Development

• Carl D. Crane III
• Professor, Department of Mechanical and Aerospace
  Engineering
• Director, Center for Intelligent Machines and
  Robotics
• University of Florida

2
Agenda for presentation

• Introduction
• Spatial Mechanism Design and Analysis
• passive and active force control mechanisms
• controlled compliance
• self-deployable tensegrity structures
• Autonomous Vehicle Development
• ground vehicle technologies
• architecture design
• micro-air vehicles

3
Introduction - Univ. of Florida

• The University of Florida is a public,
  comprehensive, land-grant, research university.
• It is one of only 17 public, land-grant
  universities that belongs to the Association of
  American Universities.
• 46,500 students
• 72 undergraduate
• 28 graduate

4
(No Transcript)
5
Introduction - College of Engineering

• 11 Departments
• 4,500 undergraduate students
• 1,900 graduate students
• degrees granted in 00-01 academic year
• 726 Bachelor
• 465 Master
• 95 Ph.D.

6
Introduction - CIMAR

• Interdisciplinary robotics research group founded
  in 1970s by Dr. Del Tesar.
• Directorship changed to Dr. Joseph Duffy in 1987.

• Currently consists of 3 faculty and 28 graduate
  students.
• 35 Ph.D. graduates and 70 M.S. graduates in
  period 1985-2000.

7
Introduction - CIMAR

• Joseph Duffy
• graduate research professor
• fellow ASME
• received ASME Machine Design Award in September
  2000 for "eminent acheivement in the field of
  machine design"

8
Agenda for presentation

• Introduction
• Spatial Mechanism Design and Analysis
• passive and active force control mechanisms
• controlled compliance
• self-deployable tensegrity structures
• Autonomous Vehicle Development
• ground vehicle technologies
• architecture design
• micro-air vehicles

9
Passive force control mechanisms

• Industrial robots are typically six degree of
  freedom devices.
• The torque of each actuator is
• commanded by controlling current.
• Closed loop joint position control
• is typically enacted by using
• optical encoders to provide
• feedback.

10
Passive force control mechanisms

• Simple applications or industrial robots are
• position control operations where the manipulator
  does not contact the environment
• spray painting
• welding

• force control operations
• where the end effector
• is constrained so it
• cannot move

11
Passive force control mechanisms

• Research objective
• develop a means whereby the end effector of a
  serial robot manipulator can be controlled for
  position and force
• Approach
• design and implement a compliant wrist mechanism

12
Planar example

• it is desired to control the contact force of a
  wheel as it moves along a rigid wall
• the wheel is connected to a simple 2 degree of
  freedom serial manipulator by a two-spring system
  the lengths of the springs are measured
• input parameters are ?d1 and ?d2, the change in
  displacement of the end effector
• output parameters are ?fx and ?fy, the change in
  force at point C

13
Planar example

• the relation between (?d1, ?d2) and (?fx, ?fy)
  can be written as

14
Planar example

• numerical example
• ?1 45, ?2 90
• k1 k2 10 lbf/in
• system is at unloaded position
• determine how to move robot end effector (?d1,
  ?d2) in order to reduce the normal contact force
  fn by some amount ?fn

15
Planar example

• for the given values,

16
Planar example

• the relationship between change in displacement
  and change in force can be written as
• inverting K gives

17
Planar example

• at this instant
• thus
• at this instant, the normal force can be adjusted
  by translating the end effector in the x direction

18
Second planar example

• control of force and moment
• a passive three-spring system is inserted in the
  wrist of the manipulator
• spring lengths are measured
• the relationship between the joint displacements
  and the wrench (force/ torque) applied to the end
  effector can be determined

19
Spatial passive force control mech.

• Develop a passive 6 d.o.f. parallel mechanism to
  use as a wrist element on an industrial
  manipulator.
• Platform is comprised of 6 S-P-S leg connectors.
• Leg lengths are measured.
• Spring constants and
• spring free lengths are
• measured experimentally
• during calibration.
• Relationship between dis-
• placement of top platform
• relative to the base and the
• applied wrench can be determined.

20
Determination of platform geometry

• The platform geometry must be such that a forward
  displacement analysis can be readily conducted.
• If the lines along each of the legs become
  linearly dependent, the platform is in a
  singularity and will collapse.
• For arbitrary loading, at the nominal home
  position, the device should be as far from a
  singularity as possible.

21
Determination of platform geometry

• The platform with the simplest forward analysis
  is the 3-3.
• Forward analysis reduces to a 4th order
  polynomial in t2.

• The problem with this geometry lies in the
  co-intersecting ball-and-socket joints.

22
Determination of platform geometry

• The 6-6 platform avoids the co-intersecting joint
  problem, but the forward analysis is complex.

23
Determination of platform geometry

• The special 6-6 platform was discovered.
• avoids the co-intersecting joint problem
• complexity of forward analysis is same as for the
  3-3 platform

Note that at the mechanism is in a
singularity for the configuration shown in the
figure.
24
Special 6-6 platform geometry
25
Determination of platform geometry

• The platform geometry must be such that a forward
  displacement analysis can be readily conducted.
• If the lines along each of the legs become
  linearly dependent, the platform is in a
  singularity and will collapse.
• For arbitrary loading, at the nominal home
  position, the device should be as far from a
  singularity as possible.

26
Determination of platform geometry

• For 3-3 platform
• optimal home position occurs when the base is
  twice the size of the top platform (b 2a) and
  the distance between the platforms is a.
• similar analysis performed for special 6-6

a
a
b
27
Hardware development

• Prototype device fabricated.

28
Agenda for presentation

• Introduction
• Spatial Mechanism Design and Analysis
• passive and active force control mechanisms
• controlled compliance
• self-deployable tensegrity structures
• Autonomous Vehicle Development
• ground vehicle technologies
• architecture design
• micro-air vehicles

29
Active force control mechanism

• development of parallel platform mechanism with
  actuated prismatic joints
• force in each leg sensed
• position and orientation of top platform
  determined from measurements of a separate
  metrology frame

30
Active force control mechanism

• system being designed to manipulate a 250 lbf
  load
• control system will aim to move the top platform
  in response to externally applied wrenches to the
  top platform

31
Active force control mechanism
32
Agenda for presentation

• Introduction
• Spatial Mechanism Design and Analysis
• passive and active force control mechanisms
• controlled compliance
• self-deployable tensegrity structures
• Autonomous Vehicle Development
• ground vehicle technologies
• architecture design
• micro-air vehicles

33
Controlled compliance

• a semi-active piezoelectric based friction damper
  is being incorporated into a leg connector to
  allow for control of motion damping in the leg

Actuator
Air Bearing
Spring
Friction Pad
34
Agenda for presentation

• Introduction
• Spatial Mechanism Design and Analysis
• passive and active force control mechanisms
• controlled compliance
• self-deployable tensegrity structures
• Autonomous Vehicle Development
• ground vehicle technologies
• architecture design
• micro-air vehicles

35
Tensegrity structures

• comprised of struts in compression and ties in
  tension

36
Self-deployable tensegrity structures

• certain ties are replaced by elastic members

37
Self-deployable tensegrity structures
38
Self-deployable tensegrity structures

• analyses conducted to
• determine deployed position at equilibrium
• determine motion of system in response to
  externally applied loads
• determine motion of system in response to change
  in spring free lengths

unloaded and final equilibrium positions
39
Application of tensegrity mechanisms

• self-deploying satellite antennae

40
Application of tensegrity mechanisms

• tents and shelters

41
Application of tensegrity mechanisms
b
a
c
d
42
Agenda for presentation

• Introduction
• Spatial Mechanism Design and Analysis
• passive and active force control mechanisms
• controlled compliance
• self-deployable tensegrity structures
• Autonomous Vehicle Development
• ground vehicle technologies
• architecture design
• micro-air vehicles

43
UF systems
44
AFRL systems
45
Autonomous vehicle technologies

• path planning
• obstacle detection and mapping
• positioning systems
• vehicle control
• system architecture

46
Autonomous vehicle technologies

• path planning

47
Autonomous vehicle technologies

• path planning
• obstacle detection and mapping

48
Autonomous vehicle technologies

• path planning
• obstacle detection and mapping
• positioning systems

49
Autonomous vehicle technologies

• vehicle control

follow-the-carrot
pure pursuit
50
Vehicle control
follow-the-carrot
pure pursuit
51
Vehicle control

• previous two methods only aim to move the vehicle
  to the goal point
• the desired orientation at the goal point is not
  being considered
• errors in position and errors in orientation have
  different units
• led to a new technique called vector pursuit
• determine a screw that will correct translational
  error
• determine a screw that will correct rotational
  error
• sum them together to get a desired instantaneous
  motion screw

52
Vector pursuit - method 1
goal pose
vehicle pose
path
53
Vector pursuit - method 1

• the desired screw must be mapped to the allowable
  motion space of the vehicle

commanded twist
desired twist
54
Vector pursuit - method 2

• take into account vehicle motion constraints when
  determining the screw to correct translation and
  rotation

path
vehicle pose
55
Fuzzy reference model learning control
56
Vector pursuit implementation
57
Autonomous vehicle technologies

• path planning
• obstacle detection and mapping
• positioning systems
• vehicle control
• system architecture

58
System architecture development

• UF is a member of the DoD Joint Architecture for
  Unmanned Ground Systems (JAUGS) Working Group
• JAUGS goals
• reduce life cycle costs
• reduce development and integration time
• provide a framework for technology insertion
• accommodate expansion of existing systems with
  new capabilities

59
JAUGS

• the architecture is comprised of a set of
  components with well defined interfaces
• the components are designed for
• vehicle platform independence
• hardware independence
• technology independence
• mission isolation

60
MAX
JAUGS
PLN
MCU
OCU
sweep planner
path seg- ment driver
vector driver
go to goal planner
way point driver
velocity state driver
MRS main
obstacle avoidance
MRS
MRS
POS
DMS
pose sensor
obstacle detection
mapping
velocity state sensor
MRS
MRS
VCU
camera
camera
camera
actuator component
sensor component
primitive driver
video switch
MRS
61
Primitive driver component
vehicle independent input message
action description
propulsive wrench resistive wrench
Primitive Driver
effect motion via vehicle actuators
62
Wrench

• set of six ordered values
• fx, fy, fz mx, my, mz
• values sent as a percentage

F
mx
x
my
y
mz
z
63
example 1 teleoperation
OCU
wrench motion commands defined in local coord.
sys.

Primitive Driver Component
vehicle motion
64
velocity state driver component
velocity sensor message
desired velocity state
Current velocity state. vehicle coordinate system
The velocity that the vehicle should have at this
instant (instantaneous motion screw). vehicle
coordinate system
Velocity State Driver Component
wrench motion command
Desired propuslive and resistive force acting on
vehicle. vehicle coordinate system
65
example 2 operator specifies velocity state
global coordinate system
OCU

desired velocity state defined in vehicle coord.
system
Velocity State Driver Component
wrench motion commands defined in vehicle coord.
sys.
Primitive Driver Component
Velocity State Sensor Component
vehicle motion
66
example 3 autonomous navigation
path defined as a series of waypoints measured in
global coordinate system
Path Segment Driver Component
current goal point defined in global coordinate
system
Way Point Driver Component
vehicle pose in global coord. system
desired velocity state defined in vehicle coord.
system
velocity state defined in vehicle coordinate
system
Velocity State Driver Component
wrench motion commands defined in vehicle coord.
sys.
Pose Sensor Component
Primitive Driver Component
Velocity State Sensor Component
vehicle motion
67
Component implementation
68
Agenda for presentation

• Introduction
• Spatial Mechanism Design and Analysis
• passive and active force control mechanisms
• controlled compliance
• self-deployable tensegrity structures
• Autonomous Vehicle Development
• ground vehicle technologies
• architecture design
• micro-air vehicles

69
micro-air vehicles

• work done by P. Ifju and M. Nechyba of the
  University of Florida

70
micro-air vehicles

• horizon tracking

71
micro-air vehicles
72
Recent Adventure

• The Search for Lewis and Clarks Iron Boat
• 10 - 14 Sep 2001
• Great Falls, Montana

73
(No Transcript)
74
Lewis and Clark
75
Lewis and Clark Results
76
Lewis and Clark
77
Lewis and Clark
78
Video Presentations
79
Conclusion

• Spatial Mechanisms
• Autonomous Ground Vehicles