Design Simulation in Support of NASA's Robotic and
Human Lunar Exploration Program
Powerpoint slides of this talk (48MB) Powerpoint slides of this talk (in HTML) PDF document of slides of this talk (27MB)
343 Soquel Ave, Suite 70, Santa Cruz CA 95602 USA
DigitalSpace Corporation has been building an open source real-time 3D collaborative design engineering and training platform called Digital Spaces (DSS) in support of NASA’s Exploration Vision. Real-time 3D simulation has reached a level of maturity where it is capable of supporting mission engineering design and operations using off-the-shelf game chipsets and open source physics and rendering technologies. This paper will illustrate several examples of state-of-the-art real-time design simulation utilizing DSS for the upcoming NASA lunar robotic and human exploration programs.I. Introduction
For several years, DigitalSpace Corporation has been building and utilizing an open-source real-time 3D collaborative engineering, design and training platform called Digital Spaces (DSS) in support of NASA’s new exploration vision. This platform has been deployed into several NASA centers and other institutions to deliver innovative applications in almost every program, ranging from ISS training to Mars exploration.II. Early
Mobile Robotic Design Simulations
In late 2005, DigitalSpace Corporation was invited to join the NASA RLEP2 (robotic lunar exploration program, surface mission) to support the design simulation of “pre-phase A” rover concepts for a planned 2011 mission to explore cold, dark traps on the Lunar south pole. Our earlier work on a real-time simulation of the Colorado School of Mines’ prototype lunar bucket wheel excavator1,2, was employed to study future In-Situ Resource Utilization (ISRU) regolith handling operations (Figures 1, 2) and paved the way for our RLEP2 participation. Funding for the effort was covered under a NASA SBIR Phase I grant. A second phase continues to support refinement of the core DSS simulation platform as well as to develop specific applications for current NASA programs.
Colorado School of Mines’ prototype lunar bucket wheel excavator.
Figure 2. DigitalSpace model of excavator operating on moon with dust simulation.
III. The RLEP2 Simulations
This section will describe our experience using DSS to create a real-time simulation of vehicles proposed during the RLEP2 surface mobility trade studies. To begin the effort, the DigitalSpace team developed a ‘moon hazards yard’ simulation environment and placed four vehicles within that environment for calibration of the real-time lunar physics model as well as to analyze the effectiveness of robotic vehicle configuration. Following the RLEP2 Mid-term Review in January of 2006, the team completed a newly rigged “V2” heavy RTG-powered six wheeled rover and placed it into an upgraded hazards yard environment, derived from the earlier work. Figures 3 through 12 below depict operation of this simulation within the newly designed moon hazard terrain, intended to test the RTG rover under variable conditions of crater wall traversal and hazard navigation. Note the physics engine input interface in the upper right hand corner of the simulation, which enabled NASA engineers to calibrate model parameters in real-time.
The V2 simulation allowed NASA engineers to calibrate, test and demonstrate elements of the design including:
· Set and calibrate physics engine parameters including gravitational force, static and dynamic friction, engine speed and torque, and damping coefficients.
· Articulate the mobility system to allow raising and lowering of the center of gravity for obstacle avoidance and instrument placement.
· Drive and steer the vehicle in the simulated lunar ‘hazards yard’ environment in real-time.
· Navigate hazards including crater walls, boulder fields, and negative hazards (small craters).
· Actively steer the view camera, using both static and ‘follow the rover’ modes.
Figure 3. New moon hazard yard with “V2” rover from RLEP2 mid-term.
Figure 4. Vehicle on “lander” before drive-off.
Figure 5. Vehicle traversing toward crater rim (note “dust” effects on wheel/surface contact).
Figure 6. Approaching crater rim.
Figure 7. Slant-traversing slope.
Figure 8. Mounting “high centering” hazard.
Figure 9. Surmounting hazard.
Figure 10. Mounting “rock” hazard.
Figure 11. Entering “negative” hazard.
Figure 12. Emerging from negative hazard.
Utilizing the hazards yard experience, the second phase of RLEP2 work began with the construction of a much larger ‘Polar Crater’ terrain (including a very low sun angle) in order to simulate the extreme conditions within Shackleton Crater at the lunar South Pole. The value of the Polar Crater simulation to the NASA team was demonstrated in real-time by giving engineers a preview of where and how thermal and lighting conditions suddenly and dramatically change during a rover descent into the shadowed areas. Figures 13-18 below present still images from the final real-time, drivable (using a joystick) physics-based simulation developed for the NASA RLEP2 team of their concept vehicle.
The Polar Crater simulation allowed NASA to calibrate, test and demonstrate elements of the design including:
· Set physics engine and vehicle mobility parameters.
· Spiral traverses of steep crater walls and the use of a bottom-mounted retractable plow to providing braking and centering.
· Navigation with LIDAR in the dark regions.
· Use of a drill and plow to sample the regolith at intervals on the crater bottom.
Figure 13. Pre-Phase A concept vehicle in simulated traverse departing lander.
Figure 14. Vehicle in spiral descent pattern seeking safe route to crater bottom.
Figure 15. Vehicle on steep slope using low center of gravity and braking plow to control descent.
Figure 16. Articulation of drive system used to clear obstacles on descent.
Figure 17. Plow used for surface regolith clearing for imaging.
Figure 18. Regolith sampling with drill.
Future work on mobile lunar robotic systems is anticipated to include:
· Improved fidelity modeling of the regolith including slip conditions on talus slopes and negative hazards (smaller craters), hardness of rock features, regolith bearing capacity, dust regimes, and a traversable terrain based on Dawes crater.
· Low fidelity thermal modeling of the vehicle based on angle to the sun on descent and the transition to the permanently shadowed regions. The thermal model could also potentially include dust effects.
· Low fidelity power systems modeling for preliminary vehicle design, including measurements of instantaneous, average and peak power, and the output of a mission power utilization profile that reflects the choices made by the ‘driver’ of the simulation during a particular run. It is anticipated that this tool will aid in the selection of the optimal power system (solar, battery or RTG).
· Low fidelity physics modeling of the plow and drill behavior (in collaboration with Honeybee Robotics).
· More complex lunar systems, including robotic and human vehicles as well as lunar habitation systems. Elements could include robotic lunar construction equipment and ISRU systems.
· The development of a multi-user capability, enabling mobile simulation agents or ‘drivers’ to participate via an Internet connection.IV. The LSAM Human Lander Simulation
In March of 2006 John Connolly, head of the human lunar lander team (Lunar Surface Access Module or LSAM) for NASA JSC received a positive recommendation from the MSFC ‘pre-Phase A’ RLEP2 team, and subsequently invited DigitalSpace to produce simulation sequences for the LSAM (Lunar Surface Access Module) during final descent and landing on a number of simulated lunar terrains. The resulting simulation is depicted in figures 19-24 below. The LSAM simulation activity focused on vehicle/terrain soft impact dynamics, and was used to evaluate single-point landings, ascent and multiple landing scenarios in variable terrain.
Figure 19. LSAM on final entry to simulated lunar surface.
Figure 20. Reaction Control System visual effects as physics thrust applied.
Figure 21. Descent seen from below.
Figure 22. Descent onto rock hazards, note low fidelity descent engine blast plume simulation.
Figure 23. Contact with hazards on landing.
Figure 24. Contact with slope hazard.
Future work in this area will target ray-casting to simulate LIDAR detection of surface hazards to avoid off-nominal landing conditions as depicted in Figures 23 and 24. Further work could include low-fidelity estimation of propellant consumption based on trajectory and landing choices, as well as the integration of the operational aspects of the use of future ISRU propellant sources. Other future goals include ISHM (Integrated Systems Health Management) simulation for abort and other scenarios for early concepts surrounding human and robotic return to the moon in the coming decade. Much will need to be drawn from the Apollo era in this work, but fundamentally real-time simulation tools will enable a new level of engineering detail to emerge for future space operations scenarios, leading to new ways of creative thinking4 for engineers and program managers.V. Brief Overview of the DSS Simulation Architecture
It has been only in the past few years that ubiquitous and powerful 3D chipsets have become standard in virtually all PCs and laptops. Therefore, real time 3D rendering is possible, with the end user able to change elements of a simulation on the fly. These environments are moving into the engineering mainstream. As an early investor in much of the serious real-time 3D industrial applications3, NASA has a great legacy to draw on for 3D in mission simulation, training and mission operations6. In the new era of exploration, NASA has a chance to again pioneer innovative uses of real-time 3D. In addition, in the era of open systems standards and open source, platforms can take advantage of tremendous economies of scale in development. With support by NASA, USRA, Raytheon and others since the year 2000, DigitalSpace has built and extended the DSS platform, which has been customized for a range of NASA mission and knowledge sharing needs.
Fig 25. Functional block feature set enabling vehicle agent simulation.
The specific extensions and enhancements to DSS are detailed in figure 25 above. The DSS Core Module binds together an agent (representing the vehicle) and the 3D scene graph (Environment), physics engine and user input devices (joystick or keyboard in this case) to produce a real-time vehicle/surface simulation.VI. Commercial Applications
In order to illustrate the value of the DSS platform to industries outside of aerospace, a prototype real-time open-pit mining simulation was developed by DigitalSpace. The results of that project are shown in Figures 26 and 27 below. The use of the DSS platform for real-time simulation applications within the mining industry has great potential. The mine simulation was developed to facilitate NASA technology transfer, an important element of the SBIR program.
Figure 26: Overhand shovel loading haulage truck
Figure 27: View of mining scene showing equipment, pit and haul road
Note that the mining simulation above represents an early stage model. Completed elements included vehicle mobility, geometric layout of the open pit and haulage ramp, automation of haul truck traverse using waypoints, and construction of the general environment.
Current capabilities of the mine simulation using DSS platform include:
- Used to visualize both open pit and underground mining operations
- Virtual models of selected mining equipment were built and assigned physics properties
- Real-time pit or underground 3D mine data could be viewed by managers or engineers with ability to steer camera view angle to any geometry desired
- Could be used as an operations planning and training tool with 3D visual feedback
Additional SBIR-funded work is currently underway to further demonstrate the capability of DSS to the mining industry. A major Canadian mining company has offered to serve as the first mining industry ‘customer’ for a custom simulation application: A real-time driveable two boom drill jumbo in a simulated 3D underground mine environment (see Figures 28 – 29 below). The model, under development at this stage, is intended to serve as a drilling operations analysis, planning and training tool. Utilizing custom developed physics parameters, it is expected to perform in real-time as a close proxy to an actual drilling unit, currently in use at the underground mine. Intended applications of the simulation include time studies for production drilling sequences and roof bolting, as well as a visual demonstration tool to describe new sequences to management and the labor force. It is anticipated that the DSS platform may emerge as an important tool for operations research and optimization.
Fig 28. DSS utilized in an underground mine to simulate a two boom jumbo drill.
Fig 29. Boomer drill making a turn.
Real-time 3D simulation using off-the-shelf game chipsets and open source physics and rendering technologies has reached a level of maturity where it is capable of supporting elements of robotic mission design. However, achieving “ground truth” in terms of a fully trustable simulation of vehicle dynamics and the physical properties of terrain (Lunar or terrestrial) is several years away. Comparison of simulation behavior with instrumented sandbox testing will be one way to partially validate the models. During future lunar surface mission operations, recalibration of the simulation model may yield a high enough fidelity in the tool that it could then be used as a predictive mission support environment for path and motion planning and day-to-day vehicle operations as is partially implemented today in the Mars Exploration Rover program3. The authors issue an open call for collaboration to government, industry and academic partners in the further development of the DSS real-time simulation platform, as well as specific lunar and planetary exploration applications. DSS lunar mission simulations can be downloaded and run on ordinary PCs with mid-grade 3D graphics capabilities. Individual simulations as well as source code are available for the platform at the web site listed in the references5.Acknowledgments
We would like to thank the following people and institutions for their support of this work: Mark Shirley (COTR), NASA Ames Research Center; Bart Graham, Al English, NASA Marshall Space Flight Center; John Connolly, NASA Johnson Space Center; and Michael Duke, Masami Nakagawa, Colorado School of Mines - Center for Space Resources.References
1Muff, T., Johnson, L., King, R., and Duke, M.B., “A Prototype Bucket Wheel Excavator for the Moon, Mars and Phobos,” Proceedings of the 2004 Space Technology and Applications International Forum (STAIF-2004),
Albuquerque, New Mexico, February 8-11, 2004.
2Damer, B. Brandt, G., Rasmussen, D., Newman, P., “Final Report for SBIR I: NNA05AC13C, Simulation-Based Lunar Telerobotics Design, Acquisition and Training Platform for Virtual Exploration,” DigitalSpace Publications, July 2005, available on the web at URL: http://www.digitalspace.com/reports/sbir04-phase1-finalreport/
3Baumgartner , E. T., “Motion Planning Technologies Planetary Rovers and Manipulators,” International Workshop on Motion in Virtual Environments, LAAS-CNRS,
Toulouse, France, January 8, 2005.
4Michael B. Duke, Stephen J. Hoffman, and Kelly Snook. “Lunar Surface Reference Missions: A Description of Human and Robotic Surface Activities,” NASA TP-2003-210793,
NASA Johnson Space Center, July, 2003.
5The DSS-Prototyper simulations described in this report are available for beta testing installation at URL: http://www.digitalspace.com/projects/lunar-robotics/index.html [cited 22 August 2006] and DSS source code (when released) can be downloaded and built from the following library URL: http://www.digitalspaces.org [cited 22 August 2006].
6R.B. Loftin and P.J. Kenney, Training the Hubble Space Telescope Flight Team, IEEE Computer Graphics and Applications, vol. 15, no. 5, pp. 31-37, Sep, 1995.
 President and CEO, DigitalSpace HQ,
343 Soquel Ave, Suite 70, Santa Cruz CA 95062.
 Director, DM3D Studios, DigitalSpace Corportaion.