Biomimetic Robots for Robust Operation in Unstructured Environments

1
Biomimetic Robots for Robust Operation in
Unstructured Environments

• M. Cutkosky and T. KennyStanford University
• R. Full and H. KazerooniU.C. Berkeley
• R. HoweHarvard University
• R. Shadmehr
• Johns Hopkins University

http//cdr.stanford.edu/touch/biomimetics
2
Behold Behemoth ... His bones are tubes of
bronze, his limbs like bars of iron.
Job 40.18
Boadicea climbing a rock, by M. Binnard
3
Main ideas

• Use novel layered prototypingmethods to create
  compliant biomimetic structures with embedded
  sensors and actuators (Cutkosky, Kenny, Full)
• Develop biomimetic actuation and control schemes
  that exploit preflexes and reflexes for robust
  locomotion and manipulation (Kazerooni, Howe,
  Shadmehr, Cutkosky)

4
Status (9.10.98)

• Building block design/fabrication environment
  being tested and first components with embedded
  sensors, electronics fabricated
• Meetings among SU, SRI, UCB to determine
  biomimetic actuators for fabrication and testing
  at each site
• Designed and built apparatus for leg stiffness
  and perturbation experiments
• ltHowe?gt
• Experimental results on human adaptive control
  suggest a specific design for manipulation
• Modeling and system I.D. have been applied to
  capture human walking on hills. The results have
  been used to develop two-legged machines.
  Comparison with biological models is underway.

5
Topics for discussion and planning

• I. Mobility (Full, Kazerooni, Howe)
• Autonomous and cooperative
• Lessons from insect and vertebrate biology
• II. Manipulation (Howe, Shadmehr, Cutkosky)
• In insects, in humans
• Role of sensing, adaptation
• III. Learning and adaptation (Shadmehr, Howe,
  Kazerooni)
• Focus on adaptation for cooperative mobility,
  manipulation
• Role of adaptation for robust autonomous robots
• IV. Fabrication integration experiments
  (Cutkosky, Full, Kenny)
• V. Sensing (Kenny, Full, Howe, Cutkosky)
• VI. Actuation and control (Cutkosky et al)

discussion leader
6
Mobility
Level of Interaction with Human
Power
PowerInformation
Information
Autonomous
Collaborative
Unpowered Lower Extremity Enhancer
Powered Lower Extremity Enhancer
Powered Human Assisted Walker
Mobile insect
7
Mobility Modeling insect locomotion dynamics
A dynamic cockroach model, created in
collaboration with the MIT Leg Lab, is stable
when stiffness and damping feedback are added to
the feed-forward joint torques (R. Full)
8
Modeling human locomotion
(Kazerooni)
9
Adaptation in manipulation
Experimental Framework for understanding how
humans go about modulating impedance while
interacting with an unstable system (Shadmehr,
Kazerooni)
10
Motor control, adaptation model
(Shadmehr)
11
Fabrication and joint experiments concept for a
biomimetic Insect-Leg
A prototype design of the same leg employing
three-dimensional plastic exoskeleton
surrounding with embedded actuators, sensor and
cooling system.
12
Sensing MicroStructures and Sensors Lab (MSSL)
Kenny

• Research on Fundamental Properties and
  Applications of MEMS-based MicroMechanical
  Devices.
• Micromechanical Sensors.
• Micromechanical Elements for Scientific and
  Technological Collaboration Partners.
• Devices and Instruments for Studies of
  Fundamental Properties of Micromechanical
  Structures.
• Collaborators IBM, JPL, NRL, SNL, SAIC,
  Medtronic, Raychem, Lucas, Seagate,
  Perkin-Elmer...
• Students from ME, EE, Appl Phys, A/A

Piezoresistive Lateral Accelerometer
2-Axis AFM Cantilevers for Surface Friction
Experiments and Thermomechanical Data Storage
Flow Visualization in Microchannels
Ultrathin Cantilevers for attoNewton Force
Detection
13
Actuation and control Mechanics and muscle
activation patterns (Full)
Three-dimensional musculo-skeletal model of the
leg of B. discoidalis constructed by Fulls lab.
Simulations such as these help characterize the
role of individual muscles in locomotion.