2.2.5 Free Flying Servicers

This segment of the program is focussed on the development of space robotics for eventual application to on-orbit satellite servicing by free-flying servicing robots. The activities targeted by these robotic systems are primarily tasks currently performed by astronauts in pressure suits. These tasks are critical to planned and contingency space operations, both on the International Space Station as well as for major science programs such as Hubble Space Telescope. The goals of this program segment are to develop technologies and integrated systems concepts that to form a more capable team of humans and robots for complex maintenance and assembly tasks, and to reduce the need for human EVA through robotic performance of simple or repetative tasks.

Required technologies to be developed under this segment of the Telerobotics Program include assisted teleoperation with time delay, virtual reality telepresence, advanced display technologies, proximity sensing for perception technologies, and robotic flaw detection. The target applications include such tasks as cooperative human-robotic interactions in an EVA work site, free-flight technologies for inspection and monitoring, and ground-based control of robotic servicers. Each of these areas have been identified by the potential space robotics user community as applications where space robotics will be necessary to satisfy their planned requirements. This user community includes the International Space Station, the advanced EVA community, External Work System, and anticipated commercial space system developers.

Ranger Telerobotic Flight Experiment

Ranger Automated Visual Inspection System

AERcam II

Free Flying Camera Robots For Enhanced EVA Performance

Technology Transfer Roadmap Details

Ranger Telerobotic Flight Experiment

In June 1992, the decision was made to actively pursue the development of the Ranger Telerobotic Flight Experiment (TFX), as proposed by the University of Maryland Space Systems Laboratory. This project included the development of neutral buoyancy and flight prototypes for a class of low-cost expendable telerobots designed for research and servicing in space, beyond the space station orbit. The vehicles were designed to incorporate four manipulators: two 8-DOF arms for bilateral dexterous manipulation; a 7-DOF manipulator for grappling at the local worksite; and a 6-DOF arm for positioning a pair of stereo video cameras giving primary feedback to the remote operator. A second stereo camera pair mounted on the vehicle centerline provides a stable visual reference for free-flight maneuvering, and ultimately feed a vision system for autonomous vehicle docking. Both the Ranger neutral buoyancy vehicle (NBV) and the space flight vehicle for the Telerobotic Flight Experiment (TFX) were designed with appropriate systems for full free-flight capability.

Development activities following the program inception focused on the fabrication, assembly, and testing of the neutral buoyancy vehicle, and the detailed design of the TFX flight vehicle. By the end of FY96, Ranger NBV had become a capable and reliable research vehicle, incorporating highly capable dexterous manipulators and nearing integration of the grappling and video manipulators. At the end of 1996, Ranger NBV participated in an extensive series of tests with EVA subjects at the NASA Marshall Neutral Buoyancy Simulator. During these tests, Ranger NBV successfully performed a number of tasks from Hubble Space Telescope servicing and International Space Station maintenance, both solo and in cooperative activities with EVA subjects. During these tests, Ranger NBV was primarily controlled via satellite from NASA Johnson Space Center and from the University of Maryland Neutral Buoyancy Research Facility.

During this same period, Ranger TFX underwent detailed systems design, culminating in the second of two detailed design reviews in April, 1996. Hardware fabrication for the space vehicle were placed on hold pending the successful completion of an extensive series of inquiries into non-NASA funding sources for approximately $20M to pay for flight on an expendible launch vehicle. Without significant success in this area, Ranger TFX was terminated at the end of FY96.

With the significant technology gains of the Ranger NBV programs, and the continued existence of the need for low-cost flight demonstration of dexterous telerobotics, the NASA/University of Maryland team significantly redesigned the TFX hardware to arrive at a functional design for a shuttle-based flight experiment. This program, called Ranger Telerobotic Shuttle Experiment (TSX), will be a joint endeavor between the University of Maryland and the NASA Johnson Space Center, aimed at making maximum use of Ranger TFX technologies and designs in implementing a low-cost, quick-development shuttle flight experiment.

Ranger TSX adopts the design of the TFX manipulator and electronics modules, along with all four manipulators proven in NBV development and testing. It will be permanently attached by the end of the grapple manipulator to a Spacelab pallet in the Space Shuttle payload bay, and released from launch restraints to move about the pallet to address a number of varying test hardware interfaces. Primary emphasis on the Ranger TSX flight experiment will be as a risk mitigation experiment for the extensive telerobotic operations planned for the International Space Station. As such, it will incorporate ISS maintenance tasks in its design test set, along with HST servicing tasks from the TFX mission design set. Current plans are for Ranger TSX to be operable from the Shuttle aft flight deck, but primary control will be from a ground station at the Johnson Space Center.

Focus and Directions:

FY 1994: Complete and test the neutral buoyancy version of Ranger; complete Phase A and B studies on the Ranger Telerobotic Flight Experiment (TFX)

FY1995 Collect engineering data from Ranger NBV in the UMd Neutral Buoyancy Research Facility; continue detailed design and long-lead item procurement for Ranger TFX

FY1996 Collect operational data from Ranger NBV for correlation with Ranger TFX flight data; construct and test Ranger TFX subsystems

FY1997 Perform mission operations training and continue operational data collection using Ranger NBV; integrate Ranger TFX systems and perform integrated test and verification of TFX vehicle in preparation for FY98 launch

FY1998 Fly Ranger TFX and collect operational data for correlation with Ranger NBV data; analyze data collected from Ranger NBV and TFX operations

Additional information and specifics associated with the focus and direction of the flight experiment can be found in the Integrated Design Review documentation.

Major Milestones:

Dec 96: Ranger TSX configuration trade study complete
Jan 97: Manifesting form 1628 submitted
Apr 97: Preliminary Design Review
Oct 97: Critical Design Review
Jun 98: Flight hardware delivery to JSC
Sep 98: JSC environmental testing complete
Oct 98: Delivery to KSC
Dec 98: Flight (planning target of STS-95, 12/3/98 launch)

Point of Contact:
Dave Akin
University Of Maryland
Space Systems Laboratory
Dept of Aerospace Engineering
College Park, MD
(301) 405-1138
dakin@ssl.umd.edu

Ranger Automated Visual Inspection System

Automated robotic inspection of space platforms such as Space Station is expected to be an important element to offload time consuming inspection activities from astronauts. Jet Propulsion Laboratory recently developed such a remote surface inspection (RSI) system demonstrated successfully on the ground in a constrained lab environment. The system, however, needs further evaluations/enhancements to be robust for practical use in space applications. The objectives of this task are to 1) evaluate and enhance the existing JPL automated surface inspection technology and 2) apply this technology to the Ranger TFX mission Phase 3 Visual Inspection Task to assess/validate the telerobotic automated inspection capability in real space operation lighting conditions. The inspection taskboards will be three-dimensional with different surface textures, and will be used for both ground-based pre-flight and actual Ranger flight tests.

The automated inspection software package will be evaluated and enhanced to be robust under 1) varying ambient sunlighting and under 2) varying image mis-registrations. At present, we will assume that only on/off controllable continuous lighting will be available for the actual Ranger flight, which will achieve about 80% of our evaluation/validation goals relative to the flight inspection test equipped with electronic shuttering/strobe lighting. The image registration problem is not affected by the absence of electronic shuttering/strobe lighting. Actual video images collected under varying sunlight angles (including shadowed or dark portions of the orbit) can still determine the quality of the automated inspection algorithm quantitatively below a certain level of sunlight illumination. Further, simulated strobe lighting can be added to the video images collected during the actual Ranger flight to evaluate the dynamic working range of the automated visual inspection algorithm. University of Maryland may later opt to add electronic shuttering/strobe lighting.

Technical Objectives, Rationale, and Mission Applicability

The objective of the Ranger Remote Inspection task is to demonstrate automated visual flaw detection in space using a video camera mounted on the end of a robot arm. This task is being performed in collaboration with the University of Maryland as part of the Ranger Telerobotics Flight Experiment. The task will employ techniques developed at JPL to: 1) compensate for varying ambient lighting conditions, 2) compensate for slight misregistrations between reference (before flaw) and comparison (after flaw) images, and 3) detect true changes between reference and comparison images.

The Ranger Remote Inspection task will also demonstrate the feasibility of a distributed automated inspection system. Flaw detection will be performed on the ground using dedicated hardware which will remain at JPL in order to reduce cost, improve reliability, and simplify maintenance and resource planning. This demonstration is a key stepping stone to low-cost and robust robotic inspection of large structures in hostile environments such as the Space Station.

Technical Approach The inspection technology developed at JPL is well suited to the Ranger application and will be adapted to the Ranger task with minimal changes. On the other hand, the architectural differences between an in-lab technology demonstration and a real-world, space-deployed system are considerable. Tradeoff studies have resulted in the conclusion that a distributed inspection system would be most reliable and cost-effective. Spaceborne hardware for inspection is virtually eliminated, while the inspection system, including specialized image processing hardware and the operator interface, will remain at JPL where it can be easily maintained.

The demonstration will be performed using a taskboard designed to exercise various components of the inspection system. Prior to launch, simulated flaws will be placed on the taskboard. Inspection will take place in Earth orbit. Video images will be transferred to JPL in near-real-time and detection of the flaws will be verified. Camera position will be controlled from JPL with ultimate authority to remain with Ranger mission control at the University of Maryland.

Focus and Directions:

FY 1996: Evaluate/enhance the existing automated inspection software package. Design a Ranger inspection board mockup and build two (one for JPL test and the other for Ranger delivery). Define onboard and ground-site requirements for the Ranger flight inspection experiment in coordination with the University of Maryland Ranger TFX team. Define inspection scenario, and develop inspection task software including user interface and command sequences consisting of arm motion control, light on/off, and image captures. Deliver a taskboard mockup, inspection task software, and inspection scenario to the University of Maryland Ranger Program for Neutral Buoyancy test by June 1996.

FY 1997: Deliver a revised inspection taskboard for the Ranger flight by December 1996. Complete the development of an enhanced automated inspection software package and perform ground baseline pre-flight tests the end of 1997. Analyze, on the ground, actual inspection experiments in the flight and post-flight operations. Document the results.

Major Milestones - FY97

Demonstrate near-real-time remote inspection of unknown flaws using images of JPL-built taskboard from Maryland.

Collaborative/Other Supporting Work

The Ranger Remote Inspection task represents an application of remote inspection technology developed at JPL under the Remote Surface Inspection (RSI) project from 1990 to 1994. The RSI task demonstrated automated ambient-light-compensated visual flaw detection using image differencing.

Most of the inspection techniques used in Ranger Remote Inspection are directly based on the results of the RSI research. The inspection platform developed for the RSI project, including a multi-modal inspection end-effector, a dedicated image processor, and operator interface software, is used to modify and test the Ranger Remote Inspection system.

Point of Contact:
Won S. Kim
Jet Propulsion Laboratory
Pasadena, CA 91109
(818) 354-5047
wonsoo@telerobotics.jpl.nasa.gov

AERcam II

(Note: In FY 1997, portions of this task were transferred to the Office of Space Flight robotics program)

Objectives:

The Autonomous EVA Robotic Camera (AERCam) is being developed to perform visual and non-visual inspection activities around orbiting parent spacecraft. The ultimate goal is to have a system that can be stored in a carrier, just like putting a car in a garage, and that can then be commanded to fly to a specific location on the parent spacecraft to perform an automated inspection. The system will autonomously fly to the commanded location, perform the inspection, and intelligently screen the data to determine if human intervention is required. Progress toward that ultimate goal will be made in three phases. In its final configuration, AERCam will have the capability to:

autonomously fly to a commanded location (such as a desired viewing position for a point of interest) while avoiding obstacles,

autonomously fly search patterns for visual or non-visual inspections,

intelligently screen inspection data to determine if further human analysis is required, and

autonomously stationkeep to an EVA astronaut to send camera views of EVA activities back to the IVA crew and ground controllers.

The AERCam project is divided into three phases.

AERCam I, called Sprint, is teleoperated with hand controllers by an IVA crew member and is scheduled to fly on STS-87 (currently manifested in October 1997). Sprint is a free flying spherical vehicle, teleoperated by an IVA operator, with a stereo camera pair to provide vision feedback to the operator. Video images from a stereo camera pair onboard the Sprint platform will be sent to the IVA crew and ground controllers. The vehicle weighs 35 pounds and is packaged into a 13 inch diameter aluminum sphere covered with Nomex felt (see Figure 2). Sprint is manifested to fly on STS-87, scheduled for October 1997, as a Phase 1 Risk Mitigation Experiment for the ISS.

AERCam II automates the guidance, navigation, control, stationkeeping, and inspection functions to meet the technical objectives. The AERCam is given task level commands by the operator through either a point-and-click graphical user interface or a voice command system. An Integrated Ground Demonstration (IGD) is planned that will demonstrate the functionality of the system near the end of FY97. The hardware of the IGD unit will then be redesigned and repackaged with an emphasis on miniaturization for a Shuttle flight experiment.

AERCam III will increase the intelligence of the system and transition much of the AERCam II off-board processing (such as vision processing) onto the free-flyer. Accordingly, miniaturization of hardware will continue to be a major focus. In addition, AERCam III will provide the ability to interface non-visual sensor payload(s) with the free-flying platform. AERCam III will focus on technologies necessary to perform autonomous inspections once at the site. The AERCam can be used by the Space Shuttle, the International Space Station (ISS), or a transfer vehicle to the Moon or Mars. It will also improve the robustness of AERCam II systems and address technical difficulties encountered with AERCam Sprint and AERCam II.

The near term objective is to develop perception, navigation, and intelligence technologies and test them as part of an Integrated Ground Demonstration. For the perception technologies, technical advances will be made to allow operation in the wide range of lighting conditions experienced in space. Additionally, advances will support recognition and tracking of objects of interest that are partially occluded. Technical advances will be made in the navigation technologies in the determination of a relative position and attitude of a spacecraft. A significant challenge to be addressed by the intelligent software is the desire to use a common architecture for both autonomous and teleoperated control modes as well as to smoothly transition between these modes. The system will be required to perform many of the operations in real-time. Data from multiple sensors will require validity checks and must be fused in a manner that resolves any conflicting information.

Approach:

The goal of AERCam II is to provide autonomous functions that free the operator from having to control the free-flyer via a hand controller. Although AERCam II will provide teleoperational capabilities as a backup to autonomy, the user will be able to issue task-level commands that the system will perform autonomously. The capabilities being developed for AERCam II will enable the system to monitor and control free flyer position and attitude with respect to the parent vehicle or with respect to an object in the field of view of the cameras. The autonomous functions will include the ability to detect and avoid obstacles. After the free flyer has autonomously flown to a desired location, it will be able to fly user-selected paths to acquire inspection images.

This section describes key AERCam II technologies and how they will be employed in the AERCam II system to meet project technical objectives. The planned Integrated Ground Demonstration is also described, along with the technical progression expected to follow in AERCam III.

There will be three components to the AERCam II user interface: a graphical user interface (GUI), voice control, and a hand controller. The operator will be able to use either the "point-and-click" GUI or voice control to issue task-level commands to AERCam from the control station. Using the GUI, the operator will be able to direct the free flyer to maneuver to an arbitrary location by selecting the corresponding map location on the control station's computer display. The voice control will support a subset of the GUI commands, allowing the operator to control the free flyer by speaking the commands and receiving verbal feedback. After selecting or stating the desired destination (using the GUI or voice control, respectively), the planned path will be displayed and the progress of the free flyer along the planned trajectory will be charted. A hand controller will be available to support teleoperation capabilities as a backup system.

The commands issued by the operator will be managed by a multi-tiered control architecture. This architecture uses the lower two layers of the 3-Tier (3T) Intelligent Software Architecture. The 3T system consists of a planner, a sequencer, and a skill manager. The planner reasons about the long term implications of actions and possible conflicts in meeting multiple goals. Since dynamic replanning is not required for the AERCam II system, this planning tier will not be implemented. The sequencer executes hierarchical procedures, called Reactive Action Packages (RAPs), which define the sequence of steps necessary to complete an operator selected task. The skill manager orchestrates the execution of the functions defined in the RAP and monitors for task completion. The skill manager communicates task completion, or failure, to the sequencer which then transitions to another RAP. The skills implemented in AERCam II include trajectory movement, discrete movement, station-keeping, teleoperation, obstacle avoidance, landmark recognition, and visual inspection.

Guidance and navigation will be performed using a combination of a Differential Carrier Phase Global Positioning System (DCP GPS), an Inertial Measurement Unit (IMU), and computer vision techniques. A path planner will define collision-free corridors from the free flyer's current location to the desired location. The guidance system will then calculate a trajectory which is bounded by these corridors of safe passage. After the operator has approved the proposed path, thrusters on the free-flyer will be operated to maneuver along the path. The relative state (position, velocity, attitude, attitude rate) of the free-flyer will be monitored by the GPS system, using a reference GPS receiver fixed to the parent vehicle in combination with a receiver on the free-flyer, each with four antennas. In addition to GPS measurements, an IMU will be used to measure free-flyer attitude rates and translational accelerations. The IMU will continuously supply measurements to the free-flyer's navigation function, allowing continued operation in the event of GPS signal occlusion or loss of GPS phase-lock. The GPS and IMU sensors can be used when flying a trajectory between two distant points, for discrete maneuvering during inspection activities, or to maintain a fixed position relative to the reference. When the free-flyer is near a desired landmark, computer vision techniques will interact with the guidance and navigation to search for views that should be visible. Optical correlation will be used to identify the landmark and provide an estimation of the free-flyer pose with respect to the landmark. Digital vision processing may be used to refine the pose information for certain landmark-based tasks, such as autonomous docking.

Sensor input to allow stationkeeping will be achieved through two separate methods. When maintaining position with respect to the structural world model, the GPS system will provide the state data necessary to achieve the desired control. In this instance, stationkeeping will be performed using the same methods as those used for autonomous navigation. If the free flyer needs to maintain position and attitude with respect to an object, such as an item under inspection, the stereo vision system will provide the necessary state information. The stereo measurements will provide relative offset and attitude, as well as relative velocity, between the free flyer and the surface, and these measurements will be provided to the control system to maintain desired platform position and attitude. A derivative of stationkeeping is object tracking, where the free flyer is expected to keep an object of interest within its field of view, but not maintain attitude to the object (i.e., track an astronaut, but not maintain attitude with respect to his or her back). In this case, the stereo vision system will still measure the object's offset, attitude, and velocity, and the control system will use the measurements to maintain the desired offset.

During free-flyer movement, obstacle detection will be performed using a combination of stereo vision and a ring of infrared sensors. The stereo vision system will use the free-flyer's stereo camera pair to detect obstacles within a spherical region of space in front of the platform. If an obstacle enters the field of view, the stereo vision will report which portions of the spherical region are occupied and the relative velocity of the detected obstacles. A network of intersecting IR emitters and detectors (grouped in sets of one emitter and two detectors) will provide a cylinder of obstacle detection coverage to augment the stereo vision system.

If an obstacle is detected, the free-flyer navigation system will act immediately to avoid the obstacle and attempt to resume the planned trajectory. To maneuver around the obstacle, the stereo vision system will identify empty space adjacent to the obstacle and will direct the free-flyer to the empty space for safe passage. If necessary, the task level controller will be invoked to determine alternative courses of action, such as station keeping to the obstacle.

After the AERCam has collected imagery, image comparison (comparing images of the same area acquired at different times) would be performed to highlight possible problem areas. Although mission-specific image comparison is not anticipated for AERCam II, some near real-time analysis will be done as part of the inspection to demonstrate the capability to modify the inspection scenario based on the results of image analysis. Images will be registered to form a mosaic, and changes subsequently detected in the segmented regions will be reported to the user.

Milestones:

May 1996 - Conducted Sprint Preliminary Design Review
May 1996 - Began fabrication of Sprint system
June 1996 - Conducted initial Sprint Critical Design Review
Mar 1997 - Complete construction of Optical Correlator IV
Mar 1997 - Select target hardware and operating systems for free flyer and control station<
Mar 1997 - Complete preliminary GN&C architecture in MATRIX
Jun 1997 - Complete pose estimation software for first target
Jun 1997 - Complete low-level station keeping image processing and image acquisition
Jun 1997 - Complete development of control station intelligence architecture
Jun 1997 - Complete development of free-flyer intelligence architecture
Jun 1997 - GN&C software available for integration into intelligence architecture
Sep 1997 - Relative motion estimation software available for integration
Sep 1997 - Acquire custom VLSI Laplacian-of-Gaussian and binary correlator chips for compact stereo vision system
Sep 1997 - Perform AERCam II Integrated Ground Demonstration mission scenario with integrated perception technologies. (Level 1 Milestone)
Sep 1999 - Perform AERCam III Integrated Ground Demonstration (Level 1 Milestone)

Collaborative/Other Supporting Work:

The application of GPS would be a collaborative effort between JSC and Stanford University. Stanford has been working with Differential Carrier Phase GPS to provide the positioning accuracy needed to control a robot. Stanford has also developed pseudo-satellites, pseudolites, to allow application of the technology indoors and in other areas where NAVSTAR constellation GPS signals are occluded.

Flight Validation Plans:

Sprint is scheduled to fly as an EVA DTO on STS-87, currently manifested in October, 1997.

Point of Contact:
Charles B. Wheelock
Johnson Space Center
Houston, TX 77058
(281) 483-6442
charles.b.wheelock1@jsc.nasa.gov

Free Flying Camera Robots For Enhanced EVA

(Note: In FY 1997, this task was transferred to the Office of Space Flight robotics program)

Many Space Station operations will occur outside of direct line-of-sight of the habitated modules. Due to cost reduction exercises, the number and placement of mounting sites for external video cameras on Space Station will be highly constrained. Since many Space Station servicing tasks will be performed EVA, this implies that highly complex EVA operations may well take place without any capability for monitoring inside the modules or on earth. These operations could be significantly enhanced if a free-flying camera platform were available to send the desired images back. This same device could also be used to increase the performance of the robotic elements of the Mobile Work Station, by providing auxiliary views to remote operators on the station or on the ground.

This effort, performed by a University of Maryland and the Johnson Space Center team, is focusing on the technologies and operations concepts necessary for the use of free-flying cameras as operational components on the Space Shuttle and International Space Station. The University of Maryland Space Systems Laboratory will begin with the existing Supplemental Camera and Mobility Platform (SCAMP) telerobotic vehicle, and will develop new versions to incorporate advaced capabilities in the neutral buoyancy environment. The JSC Automation, Robotics, and Simulation Division (AR&SD) will be developing algorithms for realistic flight trajectories, and will develop a crew interface for SCAMP, based on existing training environments

During FY96, the prototype SCAMP vehicle was used in an extensive series of tests used to better understand the utility of free-flying cameras in conjunction with EVA operations. In January and February, 1996, SCAMP was used in the JSC Weightless Environment Training Facility (WETF). As part of this process, SCAMP became the first internally-powered robotic system to be cleared through NASA safety for use in the WETF. Over the course of these tests, SCAMP went from being limited to unattended use in the WETF, to operations during SCUBA activities, to use as a camera platform during EVA crew training for STS-80. Approximately 35 JSC personnel controlled SCAMP during these tests, and several concepts were identified for future vehicle development. Also during these test series, SCAMP was evaluated for use as a crew training aid for the AERCAM/Sprint mission. Critical development needs were identified, primarily concerned with increasing the flight dynamics fidelity, which are currently under development at the SSL.

Also during FY96, SCAMP was used in conjunction with a three-week test series at the NASA Marshall Neutral Buoyancy Simulator. During these tests, SCAMP was used on a daily basis in conjuction with EVA and Ranger NBV tests. During several of the EVA tests, SCAMP was used to monitor an EVA changeout of a Hubble Space Telescope Orbital Replacement Unit (ORU). During this process, SCAMP performed close (~2-3") visual inspection of the interior of an HST instrument bay, without inadvertant contact. SCAMP also performed close inspection of the EVA subject's pressure suit, in a pattern which would have been appropriate for checking for hydrazine contamination. As a further test, the EVA subject did not correctly close one of theHST door fasteners, and SCAMP identified the failure in a post-closeout inspection task. Throughout these tests, SCAMP was routinely controlled via satellite from the University of Maryland, and in one instance from Cypress Creek High School in Orlando, Florida.

Approach:

As a result of SCAMP operations in the past year, it has become clear that there are two distinct development paths for such robots. One is to make the system specialized to simulate the flight dynamics of a space vehicle with high fidelity in the underwater environment. Such a system will be invaluable for better understanding operational implications of free-flying camera platforms, and will be important training tools for advanced flight systems such as the follow-ons to AERCAM/Sprint. The second path is to specialize the vehicle for visual monitoring of underwater operations in support of crew training, hardware development, and so forth. This system requires a control system that supports stable viewing and reduced operator overhead, but does not need space-like dynamics. The current SCAMP prototype was designed as a compromise between these requirements, and as a result does neither perfectly.

In the remaining years of this research effort, the University of Maryland will focus on the development and test operations for two new SCAMP vehicles. Already in fabrication, SCAMP Mk. II will be a high-fidelity dynamics and control simulation of free-flying camera units in space. Incorporating acoustic-based navigational data using the 3DAPS system already developed by the SSL and flight dynamics input from JSC, SCAMP Mk. II will behave to the human operator with dynamics identical to those which would be evidenced by a flight unit such as AERCAM/Sprint. This unit will be incorporate systems of interest for operational units, such as stereo cameras, supplemental zoom cameras, voice command, and local collision detection. These candidate technologies will be evaluated in operational scenarios to assess their utility and impact on overall task performance and monitoring.

Although the flight-like unit has development priority, it is also the intention of the research team to develop a third SCAMP vehicle optimized for monitoring activities in neutral buoyancy simulations. Such a system will provide a highly stable platform for video cameras, with advanced control modes allowing stationkeeping, active following, and motion to and between predefined points in the water tank. This vehicle will also be used to perform other functions currently provided by support divers, such as detailed photography. Although less "Buck Rogers" than the space-realistic vehicle, SCAMP Mk. III will serve as a research prototype for small robotic systems which could provide better real-time video and photographic coverage than that available today, at much lower recurring costs than providing hordes of scuba divers to support the extensive neutral buoyancy development and training operations that will be needed in support of International Space Station.

During this time, JSC will be developing algorithms for realistic flight trajectories, and will develop a crew interface for SCAMP, based on existing training environments (Space Shuttle and Space Station mockups, and the IGOAL virtual reality laboratory).

Focus and Directions:

FY 1997:
Demonstrate advanced SCAMP operations in the JSC Neutral Buoyancy Laboratory
Develop trajectory algorithms for predicting free-flight dynamics in the space environment
Develop and test SCAMP Mk.II as an advanced control and operations testbed, including stereo video, position sensing and tracking, and advanced thruster design. This version will be designed to maximize fidelity of flight simulation for space flight units.
Design SCAMP Mk.III vehicle, incorporating advanced automated control modes, as system optimized for photo coverage of neutral buoyancy test operations.
Further demonstrate operational potential of free-flight vehicles in support of space station operations

FY 1998:
Verify realistic flight dynamics for SCAMP Mk. II in extended test operations
Develop and test SCAMP Mk. III in neutral buoyancy at UMd and JSC
Investigate commercial spin-offs of SCAMP Mk. III to support routine neutral buoyancy operations at JSC and elsewhere.

Point of Contact:
Dave Akin
University Of Maryland
Space Systems Laboratory
Dept of Aerospace Engineering
College Park, MD
(301) 405-1138
dakin@ssl.umd.edu