Tuesday, December 21, 2010

Robot fliers racing to catch the Zephyr

The Pentagon's hope of having a squadron of unmanned aerial vehicles (UAV) capable of staying in the air and performing surveillance for years rather than hours recently took a small step forward. Working with U.K.-based idea factory QinetiQ Group PLC, researchers from the U.S. Defense Advanced Research Project Agency (DARPA) managed to keep the solar-powered Zephyr high-altitude, long-endurance aircraft in the air over the Arizona desert for 82 hours 37 minutes, beating the 54-hour flight completed last year by an earlier version of the aircraft, the company reports on its Web site.

Launched by hand, the 66-pound (30-kilogram) Zephyr is an ultra-lightweight carbon-fiber aircraft that during the day flies on solar power generated by silicon solar arrays no thicker than sheets of paper that cover the aircraft's wings. At night it's powered by SION Power Inc.'s rechargeable lithium-sulfur batteries (recharged during the day using solar power), according to the company.

The flight, took place between July 28 and 31 as researchers guided the Zephyr by remote control to an operating altitude in excess of 60,000 ft (18 km), according toBBC News. After that, the aircraft, which carried a 4.4-pound (2-kilogram) payload, flew on autopilot and via satellite communication.

The Zephyr could be the predecessor of DARPA's proposed Vulture (Very high altitude, Ultraendurance, Loitering Theater Unmanned Reconnaissance Element) project, to create an aircraft that can remain above a surveillance target for at least five years. Under the current plans, the Vulture would weigh 1,000 pounds (453.5 kilograms) and be designed to collect its power from its environment—via solar or some other source—to store and use energy efficiently, and include a robotic refueling capability. With a wingspan of between 300 and 500 feet (between 91.4 and 152.4 meters), the Vulture would function like a low-orbit satellite as much as like an aircraft, staying aloft far longer than any surveillance plane can today.
Government contractors Aurora Flight Sciences, Boeing and Lockheed Martin are working on the first phase of the Vulture project, which began in April. The contractors are studying possible designs and configurations during this year-long first phase, which will conclude with a review of smaller and full-size demo aircraft, says DARPA spokeswoman Jan Walker. The next phase is expected to include an uninterrupted, three-month flight test of a smaller version of the Vulture.  In the third and final phase, the program will conduct a flight test of a full-scale Vulture during which the vehicle will be in the air for an entire year without landing.

"The Zephyr flight demonstrates the continuous improvement of all of the systems--solar cells, battery, light weight structural components, aerodynamic design, power management, and energy harvesting--to make (long-term, unmanned flight) possible," Jamey Jacob, an Oklahoma State University associate professor of mechanical and aerospace engineering, told ScientificAmerican.com. "I anticipate that records will begin to fall rather quickly as more platforms make use of the evolving technology."

One record that's already fallen is the current official world record for unmanned flight set by the U.S. robot plane Global Hawk—of 30 hours, 24 minutes. (QinetiQ's flight is an unofficial record because it did not involve the FAI (Federation Aeronautique Internationale), the world air sports federation, which sanctions all record attempts.

Another budding UAV project is Oklahoma State University's Pterosoar-B, which on July 1 set two new aviation world records for flight in the FAI category for autonomous aircraft of less than 5 kg with a duration of 6 hours, 15 minutes, and 54 seconds, and distance in a closed course of 122 km.

NASA Recreates Mars Surface to Liberate Rover

PASADENA – Getting stuck is never fun, especially when you’re over 30 million miles from Earth. NASA’s Spirit rover is mired in dirt on Mars and now scientists at the Jet Propulsion Laboratory are working hard to free the over-worked robot.
Spirit first ran afoul of the Martian surface on May 6 when it hit some patches of dirt that made its wheels spin in place. Now the wheels (two of which are not working properly) are sunk in up to their hubcaps.
Like a remote Auto Club for robots, JPL engineers have built a sandbox filled with a mixture of materials that closely mimic the consistency of Martian soil as well as a rock to high-center the rover. They’ve driven a replica of the Spirit into the box and are working diligently to figure out the best way to escape the talcum-like trap – a technique used with Spirit’s twin rover, Opportunity, back in 2005 when it also became stuck.
The first Mars Exploration Rover landed on the red planet in January 2004. Initially, the mission was supposed to last 90 Martian days, but Spirit exceeded that by over 20 times. Thanks to a recent dust storm, the fine dust that coated Spirit’s solar panels was blown off and it has been operating at full power for months now. If this latest obstacle can be overcome, Spirit can keep exploring even longer.
Read on to see how the JPL scientists created a little piece of Mars on Earth and get up close and personal with Spirit’s predicament.
Above: A JPL technician attaches a grounding strap to the rover before measuring the distance it traveled during the previous move. Below:
A Discovery Channel Canada film crew films the engineers as they work to get the rover unstuck.

Design a New Robotic Muscle Suit

Click image for larger.Click image for larger.
Students at Tokyo's University of Science have developed a new version of their muscle suit, a wearable robotic suit that assists the muscles when carrying out strenuous tasks.
The original version of the suit, which has been in production for several years, provides assistance to the arms and back but the new version provides assistance to the back only. That means it is lighter and more compact than the original model.
In a demonstration on Wednesday at the International Robot Exhibition in Tokyo, a student wearing the suit was able to bend down and lift 15 kilograms of weights with the assistance of the robotic suit. Doing so without assistance would be difficult for many people and could cause injury to some.
The university is still developing the suit and the model demonstrated on Wednesday was the first prototype. A production version is due some time in 2010.
With its greater assistance the original version of the suit will remain the most useful for heavier tasks.
In a demonstration of that model on Wednesday a student was asked to carry 10-kilogram bags of rice. With the suit switched off he could manage up to three bags before they started to get too heavy to carry, but with the suit switched on another two bags could be loaded into his arms. He quickly dropped the bags when the suit was switched off as without assistance it was too much weight to carry.
Such suits are being developed with an eye on assisting the physically challenged and workers carrying out physically demanding jobs.
Earlier this year Toyota Motor unveiled similar robot-assisted suits and has been testing them at factories in Japan with workers who have to lift large or heavy sheets of metal or car parts.

NASA Robot Solves 19-Year-Old Murder Mystery

NASA Robot Solves 19-Year-Old Murder Mystery
Dawn Sanchez was last seen alive when she stepped into Bernado Bass' car in 1991. Her disappearance and death remained unsolved until recently when—thanks to a little NASA robot—her murderer was sentenced to six years in prison.

Bass was Sanchez's boyfriend at the time of her disappearance and there were witness reports claiming that he shot the girl "in a vacant lot after the two got into a fight." The only problem was that no evidence to support this explanation was anywhere to be found. No car. No gun. No body.
This meant that Bass got away with the murder until recently when parts from the suspect's car were found buried in a large abandoned lot. They most likely would not have been found without the aid of the NASA equipment borrowed for the investigation. Using this equipment, investigators were able to figure out just where they needed to excavate:
The case was dismissed in 1991 due to lack of evidence. The case was recently reopened, when an informant reported that the car may have been disassembled and buried in a large abandoned lot in Alviso. The exact location in the lot was not specified, and the cost to excavate the entire area was too high. Further, the lot contained a substantial amount of buried and surface metallic debris, making a simple survey with metal detectors insufficient.
[T]he mixed team of scientists and engineers from CMIL, NASA Ames and the USGS deployed an instrumented Senseta MAX 5.0A rover hosting the research technologies under development, and mapped the magnetic environment of the survey area. The USGS received the processed data set, and after further post-processing, presented the county DA's office with their analysis and possible locations for excavation. Based on this data, the county excavated the site and retrieved car parts that matched the suspect's car.

IMPASS: Intelligent Robot with Mobility Platform with Active Spoke System

IMPASS (Intelligent Mobility Platform with Active Spoke System) is a novel high mobility locomotion platform for unmanned systems in unstructured environments. Utilizing rimless wheels with individually actuated spokes, it can follow the contour of uneven surfaces like tracks and step over large obstacles like legged vehicles while retaining the simplicity of wheels. Since it lacks the complexity of legs and has a large effective (wheel) diameter, this highly adaptive system can move over extreme terrain with ease while maintaining respectable travel speeds, and thus has great potential for search-and-rescue missions, scientific exploration, and anti-terror response applications.

Intelligent Mobility Platform with Active Spoke System
Principal InvestigatorDr. Dennis Hong
ResearchersPing Ren
Ya Wang
Blake Jeans
Mechanical Details
ControllersAllMotion EZSV10
AllMotion EZSV23
SensorsFoot Touch Sensors (12x)

Goals and Objectives

  • Classification for topology structures of IMPASS based on different ground contact points
  • Mobility analysis for different configuration cases, using both conventional and screw-based modified Grubler and Katzbachcriterion
  • Inverse and forward position analysis for the critical topology scheme of IMPASS
  • Singularity configuration identify and investigation using screw theory
  • Screw-based Jacobian analysis
  • Develop 2D and 3D motion planning strategies in unstructured terrain for both terrain sensing and non-terrain sensing configurations
  • Verify motion planning strategies in simulation and experimentally
  • Advance the capabilities of the hardware platform, including a moving center of gravity, onboard computer and power, and rugged body and components
  • Develop accurate and dependable perception units for terrain sensing and object recognition, including laser range finders and cameras


  1. Laney, D. and Hong, D.W.,”Kinematic Analysis of a Novel Rimless Wheel with Independently Actuated Spokes”, 29th ASME Mechanisms and Robotics Conference, Long Beach, California, September 24-28, 2005.
  2. Hong. D.W. and Laney, D., “Preliminary Design and Kinematic Analysis of  a Mobility Platform with Two Actuated Spoke Wheels”, US-Korea Conference on Science, Technology and Entrepreneurship (UKC 2006), Mechanical Engineering & Robotics Symposium, Teaneck, New Jersey, August 10-13, 2006.
  3. Laney, D. and Hong, D.W., “Three-Dimensional Kinematic Analysis of the Actuated Spoke Wheel Robot”. 30th ASME Mechanisms and Robotics Conference, Philadelphia, Pennsylvania, September 10-13, 2006.
  4. Wang, Y., Ren, P., Hong, D.W." Mobility and geometrical analysis of a twoactuated spoke wheel robot modeled as a mechanism with variable topology",32ndASME Mechanisms and Robotics Conference, August 6-9, 2008, Brooklyn, New York,United States
  5. Ren, P., Wang, Y., Hong, D.W." Three-dimensional Kinematic Analysis of a TwoActuated Spoke Wheel Robot Based on its Equivalency to a Serial Manipulator",32ndASME Mechanisms and Robotics Conference,August 6-9, 2008, Brooklyn, New York,United States
  6. Wang, Y., Ren, P., Hong, D.W." Gait and Gait Transition for a Robot with TwoActuated Spoke Wheels",33rd ASME Mechanisms and Robotics Conference, August30-September 2, 2009,San Diego, California, United States
  7. 2005 ASME Freudenstein/General Motors Young Investigator Award

RAPHaEL: Robotic Air Powered Hand with Elastic Ligaments

RAPHaEL (Robotic-Air Powered Hand with Elastic Ligaments) is a humanoid robotic hand that utilizes corrugated tube actuation with compressed air. Unlike electromechanically actuated hands, thanks to the natural compliance, RAPHaEL can mimic the grasping capability of a human hand more accurately. By changing the pressure of the compressed air, the amount of applied force can also be controlled.
Robotic Air Powered Hand with Elastic Ligaments
Current ResearchersKyle Cothern
Dimensions90x300mm (approx)
Weight3-5 lb (approx)
ActuatorsCorrugated Tubing
ComputingNational Instruments cRIO, custom board for mobile applications
SensorsFlex sensors for Position, Force Sensitive Resistors for Force
Power12.0 VDC, 50-120PSIG

Goals & Objectives
The goal of this project is to accurately emulate the motion and dexterity of a human hand. Once this has been accomplished, it can serve as a platform for research related to the the sensing and control in humanoid robotics. Future versions of the hand will be fabricated using SDM (Shape Deposition Manufacturing) technology which will enable us to embed all of the components into a single cohesive unit.


  • Natural compliance
  • Low in cost
  • Modular design allows for simple repair
  • Actuation requires minimul air input (15ml of air at 60psig to achieve actuation in one finger)

Potential Applications

  • Hazardous environment operation
  • Industry settings
  • Prosthetics
  • Telepresence
  • Entertainment
  • Sign language interpretation

DARwIn: Dynamic Anthropomorphic Robot with Intelligence

DARwIn (Dynamic Anthropomorphic Robot with Intelligence) is a family of fully autonomous humanoid robots capable of bipedal walking and performing human like motions. Developed at the Robotics & Mechanisms Laboratory (RoMeLa) at Virginia Tech, DARwIn is a research platform for studying robot locomotion and autonomous behaviors, and also the base platform for Virginia Tech’s entry to the RoboCup competition.

Dynamic Anthropomorphic Robot with Intelligence
Principal InvestigatorDr. Dennis Hong
ResearchersJeakweon Han
Dr. Bohee Lee
Seungmoon Song
Robert Nguyen
Michael Hopkins
Height55 cm
Weight3.9 kg
MotorsRobotis Dynamixel RX-28, RX-64, EX-106
ComputingGumstix Verdex Pro XL6P
Analog Devices Blackfin BF561
VisionVT-Cam system with dual HDR camers
PowerDual 7.4v 2000mAh LiPo batteries

Goals & Objectives

DARwIn is a research platform for studying robot locomotion and is also the base platform for Virginia Tech's entry to the RoboCup competition. In this research, we study the issues of mechanical design, kinematics, dynamic bipedal gaits, ZMP control, vision tracking, and complex autonomous behaviors needed for playing soccer.

Current Areas of Research

  • Particle-based Localization
  • Bipedal Locomotion

DARwIn Family

Many versions of DARwIn have been developed, each an improvement on its predecessor.

DARwIn 0

To investigate the feasibility of controlling a 21 DOF humanoid robot, the Cycloid robot which was designed and fabricated by Robotis was used as the testing platform. Since this was not the first physical iteration of DARwIn this testing iteration is called DARwIn 0. The motors used for controlling the robot's motion were the Dynamixel DX-117.
DARwIn 0 proved to be a success, demonstrating that the core software developed for the robot worked for controlling the robot to stand up and walk.


DARwIn I was the first humanoid robot created by a senior design project at RoMeLa. DARwIn I has 21 degrees of freedom, 4 force sensors on each feet (which were never used), a 3 axis rate gyro, a 3 axis accelerometer, and space to house a computer and batteries for powering the motors, sensors, and computing equipment. DARwIn I's links are fabricated out of bent sheet aluminum. This robot also uses Robotis Dynamixel DX-117 motors for the joints.
No sensor information was used for stability control and the gaits were not generated from mathematical functions. Therefore the robot would not walk successfully unless a sufficient sequence of stances was recorded. Additionally, the robot would fall over in the presence of any external disturbances.


DARwIn II was designed and fabricated by the 2006-2007 senior design team. Two versions were created, DARwIn IIa in the fall and DARwIn IIb in the spring. DARwIn IIa builds on its predecessor with improved mechanical design, more sensors, and added intelligence. The links were cut out of solid blocks of aluminum on a CNC mill to maximize stiffness and reduce weight.
DARwIn IIb is based on the design of DARwIn IIa, but with improvements in all categories. The motors used for articulating DARwIn's joints were replaced with a motor with twice the torque. DARwIn's link design was further refined to create even lighter weight parts. The entire computer, sensors, electronics package, and computer ports were mounted to a custom designed heat sink as a single module. This module is attached to the robot body using shock mounts, which allows easy access and removal while protecting the equipment from shock when falling.


DARwIn III was an attempt to perfect the designs before it. Successful in many ways, DARwIn III improved in a few key areas. RX-64 motors were used for almost the entire lower body in order to gain more power. RX-28 motors were used for the yawing motor in the hip to reduce the height of the robot. An RX-64 motor was used in the waist to add strength.


DARwIn IV was a radical change from DARwIn III. The decision was made to move away from the PC/104-Plus platform to a Gumstix, thereby allowing us to reduce the size and weight of the robot. 


Journal Papers

  1. Hurdus, J., Hong, D., “The Use of Hierarchical State Machines for Behavioral Programming in the DARPA Urban Challenge and RoboCup”, Springer-Verlag of Lecture Notes in Electrical Engineering (LNEE), 2009 (in print)

Refereed Conference Papers

  1. Muecke, M. and Hong, D. W., “Investigation of an Analytical Motion Filter for Humanoid Robots”, 5th International Conference on Ubiquitous Robots and Ambient Intelligence, Seoul, S. Korea, November 20-22, 2008
  2. Hurdus, J., Hong, D., “The Use of Hierarchical State Machines for Behavioral Programming in the DARPA Urban Challenge and RoboCup”, IEEE International Conference on Multisensor Fusion and Integration for Intelligent Systems, Seoul, Korea, August 20-22, 2008
  3. Muecke, M. and Hong, D. W., “The Synergistic Combination of Research, Education, and International Robot Competitions Through the Development of a Humanoid Robot”, 32nd ASME Mechanisms and Robotics Conference, New York City, NY, August 3-6, 2008
  4. Muecke, M., Hong, D. W., and Lim, S., “Precision Circular Walking of Bipedal Robots”, 32nd ASME Mechanisms and Robotics Conference, New York City, NY, August 3-6, 2008
  5. Muecke, K., and Hong, D. W., “DARwIn’s Evolution: Development of a Humanoid Robot”, 2007 IEEE International Conference on Intelligent Robotics and Systems, San Diego, CA, October 29-November 2, 2007
  6. Terpenny, J., Dancey, C., Goff, R., Nelson, D., Ellis, M., and Hong, D. W., “Success Strategies for Capstone Design Courses with Large Classes, Diverse Project Types, Small to Large Student Teams, and Varied Faculty Interests and Approaches”, 2007 ASEE Annual Conference & Exposition, Honolulu, Hawaii, June 24-27, 2007
  7. Hong, D. W., “Biologically Inspired Locomotion Strategies: Novel Ground Mobile Robots at RoMeLa”, 3rd International Conference on Ubiquitous Robots and Ambient Intelligence, Seoul, S. Korea, October 15-17, 2006

Non-Refereed Conference Papers

  1. Muecke, K. and Hong, D. W., “Development of a Fully Autonomous Humanoid Robot for Novel Locomotion Research and as the First US Humanoid Entry to Robocup”, NI Week, Worldwide Virtual Instrumentation Conference, Austin, Texas, August 7-9, 2007 (Most Outstanding Application of Virtual Instrumentation, Editor’s Choice Award Winner for Best Application of Virtual Instrumentation, Best Application of Virtual Instrumentation, Mechatronics Category Winner)
  2. Muecke, K. and Hong, D. W., “Development of an Open Humanoid Robot Platform for Research and Autonomous Soccer Playing”, 22nd AAAI Conference on Artificial Intelligence, Vancouver, BC, Canada, July 2007 (Technical Innovation Award, Judges’ Award for Mechanism Design)
  3. Muecke, K. and Hong, D., “A Reactive Approach to Behavior Based Control of a Soccer Playing Humanoid Robot”, US-Korea Conference on Science, Technology and Entrepreneurship (UKC2007), Mechanical Engineering & Robotics Symposium, Washington DC, August 9-12, 2007

Other Selected Publications

  1. Muecke, K. and Hong, D. W., “PC/104-Plus: The Brains Behind the DARwIn Humanoid Robot”, PC/104 and Small Form Factors, Journal of Modular Embedded Design, Vol. 12, No. 3, Summer 2008, pp. 26-30
  2. Muecke, K. and Hong, D. W., “DARwIn’s Evolution: Development of a Humanoid Robot for Research and Education”, Industrial Embedded Systems, OpenSystems Publishing, December 2007
  3. Hong, D., Muecke, K., Mayo, R., Hurdus, J., and Pullins, B., “DARwIn’s Fist Soccer Tournament: America’s First Entry to the Humanoid Division of RoboCup”, Servo Magazine, Vol. 5, No. 9, September, 2007
  4. Muecke, K., Mayo, R., Hong, D. W., “DARwIn: Dynamic Anthropomorphic Robot with Intelligence, Part 3 – DARwIn 2.0: The Next Generation”, Servo Magazine, Vol. 5, No. 2, February, 2007
  5. Muecke, K., Cox, P., Hong, D. W., “DARwIn: Dynamic Anthropomorphic Robot with Intelligence, Part 2 – Parts, Wires and Motors”, Servo Magazine, Vol. 5, No. 1, January, 2007
  6. Muecke, K., Cox, P., Hong, D. W., “DARwIn: Dynamic Anthropomorphic Robot with Intelligence, Part 1 – Concept & General Overview”, Servo Magazine, Vol. 4, No. 12, December, 2006