Terrestrial and Commercial Applications: This element of the program provides a means for the test and demonstration of space-targetted developed technologies in realistic operational test environments. These tasks are intended to move the technologies developed in the other elements of the program from the laboratory setting into operational use, and take advantage of the relatively easy access, well understood environments, and myriad problems to be solved to demonstrate the applicability of telerobotics. In addition, this element of the program includes tasks intended to rapidly move program-developed technology out into the commercial applications community. The intent of these tasks is ultimately to improve the national economic competitiveness of the United States and to improve the technology transfer efforts of the agency through the development of commercializable applications which draw upon space telerobotics technologies. These projects are jointly conducted by program laboratories and industrial partners to create and demonstrate full system prototype solutions to well understood terrestrial problems which can positively impact significant areas of the national economy.
Robotics Engineering Consortium
Integrated Robot
Control Architecture
Medical Applications
of MicroTelerobotics
Robotics Education
Robotics Engineering Consortium
Demeter
Agricultural Robot
Agriculture is a ripe, relatively unexploited application opportunity with uncommon advantages for commercializing mobile robotics technology. Over a billion tractor miles are driven annually, repeatedly over the same ground. Speeds are low, and precision is moderate. The terrain is mild, and proven navigation techniques apply. The goal of this CMU task is to develop, demonstrate and productize marketworthy controllers, positioners, safeguards, and task software specialized to the needs and constraints of commercial agriculture and related industries. Component technology results will be integrated onto a commercial agricultural harvester, and demonstrations will be conducted of automatically controlled harvesting operations to market relevant standards.
Point of Contact:
Dave Pahnos
Carnegie Mellon University
Pittsburgh, PA 15201
(412) 268-7084
dpahnos@frc2.frc.ri.cmu.edu
The objectives of this task are to develop a robot control architecture that combines the best of real-time and task-level control architectures, to provide graphical design and analysis tools to reduce the time needed to develop complex, distributed robotic systems, and to investigate the commercialization possibilities of such an architecture.
Software architectures provide frameworks for developing complex systems. They provide commonly needed control constructs that help structure the design problem and facilitate implementation and debugging. Robot architectures that have been developed in recent years are becoming more sophisticated and useful. This task will dramatically increase the power of existing robot architectures by producing an easy-to-use, integrated architecture that combines both task-level and real-time control.
The Task Control Architecture (TCA) and ControlShell (developed by RTI) are two architectures that have found wide acceptance in both the NASA and robotics research communities as emerging standards for control of autonomous and teleoperated robots. They provide largely complementary capabilities: TCA focuses on task-level control issues (planning, execution monitoring, and error recovery), while ControlShell focuses on real-time control issues (servo loops and state machines). The main focus of this task is to combine the two architectures, and to demonstrate the advantages of such an integrated architecture in developing complex robotic applications.
Besides the obvious interest within the NASA telerobotics community, such an integrated architecture holds great promise in areas such as Spacecraft Autonomy (New Millennium Program), factory automation and flexible manufacturing, and hazardous waste cleanup (DOE).
Technical Approach
We are adopting the RTI philosophy of "separable services," in the which the architecture provides independent capabilities that can be utilized by application developers, as needed. In this view, the task management facilities of TCA becomes a "service" that can be used by itself, or in conjunction with the other RTI services (ControlShell, NDDS interprocess communications).
To make this happen, the task management functions of TCA must be separated from its communications layer. This task will develop a stand-alone "Task Control Management Library" (TCML) of commonly needed task management functions, including hierarchical task decomposition, task sequencing, monitoring and exception handling. Integration with ControlShell will be in terms of sending signals between the TCML task trees and the ControlShell state machines. In a distributed environment, sending signals will be done by a message passing protocol, implemented using RTI's NDDS.
A concern with having such a sophisticated is that it may prove unusable. Our approach to this is to develop design and analysis tools that make it easier to build applications. RTI already provides many such tools for their own products, and we now have some experience, through the New Millennium Program, of designing tools visualizing task-level control flow. This task will develop graphical tools that will enable users to design task decomposition, execution monitoring and error recovery strategies, and then will produce code that uses TCML to implement those strategies. Similarly, we will adapt the visualization tools we have already developed to be used in the context of this new architecture.
Finally, this task will demonstrate the advantages of our integrated robot control architecture by working with some other project to use the architecture and our tools to develop a complex application. Potential applications include autonomous servicing robots, automated manufacturing, and flexible manufacturing
Major Milestones
FY 1997:
Task Management and Communications Layers of TCA
Separated - Q3
Stand-Alone Task Management Control Library Developed
- Q3
Integration of ControlShell and TCML Via NDDS -
Q4
FY 1998:
Demonstration of TCML Analysis Tool - Q1
Demonstration of TCML Design Tool - Q3,
Demonstration of Application Developed Using Integrated
Architecture - Q4
Point of Contact:
Reid Simmons
Carnegie Mellon University
Pittsburgh, PA 15201
(412) 268-2621
reids@cs.cmu.edu
Medical Applications of MicroTeleroboticsThis task is a cooperative NASA-Industry effort to develop a telerobot for breakthrough micro-and-minimally invasive surgeries. The industry partner is MicroDexterity Systems, Inc. (MDS). The Robot Assisted Microsurgery (RAMS) workstation developed in this task will allow a broader group of practitioners to successfully deliver needed healthcare services at reduced cost and will enable new procedures beyond human manual dexterity with computer enhanced controls. Surgical applications include the eye, ear, nose, throat, face, hand, spine, and brain.
The RAMS telerobotic system is lightweight, compact, extremely precise and robust - qualities that make it also ideal as a tool for flight science or as a manipulator for robotic planetary exploration.
In FY `94, the slave arm part of the telerobot system was developed. It has six degrees of freedom and positions its tip with a relative accuracy of better than 20 microns. Joint and Cartesian frame referenced control of the robot was implemented allowing demonstration of the robot in a number of modes of control.
In FY `95, the master arm part of the telerobot system was developed and integrated with the slave robot. The master arm has six degrees of freedom sensing of the operator's hand position to a relative accuracy of 30 microns and the ability to apply 2.5 Newtons of force to the operator's hand in three degrees of freedom. Cartesian frame referenced operation of the master-slave system and feedback of forces to the surgeon's hand simulating dynamic environments were demonstrated. A meeting of the Medical Advisory Board, composed of prominent surgeons and healthcare industry professionals and formed to guide the RAMS development, was convened in Nov. 1994.
In FY `96, detail design documentation of the slave arm was delivered to MDS to facilitate transfer of technology for commercialization. A second and improved prototype master-slave system was also developed. Upgrades were made on the first prototype slave robot for delivery to the Cleveland Clinic as recommended by the RAMS Medical Advisory Board. Upgrades included the mounting of a surgical instrument on the first prototype slave robot and the use of a joystick input device for controlling the robot in a Cartesian referenced coordinate frame. The second prototype system was also used in the demonstration of an eye microsurgery procedure - the removal of a micron-sized particle from a mock-up of a simulated eyeball. A micro-forceps surgical instrument was mounted on the slave robot and indexed control and a pivoting algorithm were implemented to enable execution of the demonstration.
In FY `97, the RAMS task will add force reflection to the RAMS system. Forces felt at the surgical instrument tip will be amplified 2-5 times at the operator's hand. Independent evaluation of the RAMS system will continue at an external lab. and operator performance experiments in using the RAMS system will be conducted. A dual arm telerobotic demonstration using the first and second prototypes of the RAMS system to perform a simulation of microsurgery suturing is planned for end of FY 97.
Technical Approach
NASA-JPL's role is to design, fabricate, test, and document an engineering R&D prototype. There is close cooperation between the JPL engineering team and MDS for determining specifications for the design. Extensive use of vended components is made to reduce development cycle time and insure reliability of the final product. Innovative mechanical, electronic and control systems design and algorithm development are utilized when necessary to achieve the unique specifications of this workstation.
MDS, working through independent funding, defines medical requirements, collaborates in system engineering design, develops and tests a pre-commercial clinical prototype, and with end market support certifies and commercializes a final design.
Review of progress and guidance on the technology commercialization path is provided by the Medical Advisory Board. Periodic meetings of the board at JPL provide an independent assessment of the technology by experts from the surgical and medical communities.
The approach for FY `97 is to build on the telerobotic platform already developed and to utilize the existing collaborative agreements. JPL will work with the vendor of the force sensor to modify their product for the RAMS application. Software development will be implemented within the Control Shell/VxWorks real-time computing environment upgraded with faster processors and a hard disk. External testing of the telerobotic system will be conducted at the Neurosurgery laboratory of the Manhattan VA Medical Center and operator performance tests will be conducted with the assistance of MDS and local medical schools.
Focus and Direction
FY `94 Develop and demonstrate slave arm
FY `95 Develop master arm and integrate with slave
robot.
FY `96 Demonstrate the telerobotic system performing
a simulated eye microsurgery procedure.; Conduct external tests at the Cleveland
Clinic Foundation.
FY `97 Implement amplified force reflection.; Conduct
external tests at the Manhattan VA Medical Center.; Evaluate performance
of the telerobotic system; Provide detail documentation to MDS; Demonstrate
a dual-arm manual telerobotic surgical suturing simulation.
Major Milestones
Level 1:
Sep 97 - Demonstrate dual arm force reflecting
teleoperated microsurgery suturing. Technical challenges are: stable control
of 2-5 times amplified force feedback; telemanipulation; combined surgical
instrument actuation and simultaneous high fidelity force sensing; operator
training for complex dual arm task execution
Level 2:
Jan 97 - Complete upgrade of 1st prototype master-slave
system and make ready for loan to testing facility
Jan 97 - Complete force sensor installation and
associated electronics on slave robot
Mar 97 - Complete software development for force
feedback
Jun 97 - Complete software development for force
and shared control capabilities
Aug 97 - Complete performance experiments on telerobotic
system
Sep 97 - Complete documentation on master-slave
system and deliver to industry partner
Collaborative/Supporting Work:
This task is collaborating with MDS under a Technology Cooperation Agreement.
MDS was formed by Steve Charles, MD, a world renowned vitreoretinal surgeon,
to develop a telerobotic workstation for microsurgery. MDS will be building
a second generation telerobotic system using over $2M in independent funding
from personal and private investment and additional funding from the healthcare
industry and government. MDS has provided the development effort at JPL
with design specifications necessary for a commercially viable product and
plans to perform clinical tests and seek FDA approval for the commercial
version of this device.
Point of Contact:
Hari Das
Jet Propulsion Laboratory
Pasadena, CA 91109
(818) 354-9174
hari@telerobotics.jpl.nasa.gov
Point of Contact:
Dave Lavery
(202) 358-4684
dave.lavery@hq.nasa.gov
with any comments, criticisms
or corrections for this page.