Terrain Relative Navigation (TRN) sensor: Astrobotic's Optical Precision Autonomous Landing (OPAL) sensor¶
Peregrine Mission One · failed to land · commercial payload
A stand-alone, passive optical navigation sensor (about 2.5 kg, 15 W peak): a camera and processor that match descent images to onboard maps made from orbital imagery, to estimate the lander's position and velocity. On Peregrine it was to estimate and log the lander's position from the start of powered descent to terminal descent, a step toward landings within 100 m.
PI: Andrew Horchler (Astrobotic)
Built by: Astrobotic Technology, with Moog (avionics and structure), NASA JPL (flight-proven FPGA vision modules, algorithm advice) and NASA Johnson (hardware-in-the-loop testing) (company)
| Technology | landing sensor (terrain relative navigation) |
| Moon Base need | #1101 (The study's reading: gap #1101 and the Users Guide's landing challenge do not name TRN, but they ask for precise, safe landings. A passive camera-based sensor like OPAL needs lit terrain; the guide's challenge stresses low-visibility terrain and shadow, which OPAL does not cover.) |
| Route onto CLPS | STMD; Astrobotic's own technology, flown on its own lander as part of an STMD Tipping Point (Game Changing Development) that began in May 2019 and required a lunar flight demonstration. Wikipedia lists it among Peregrine's payloads; Astrobotic's post-mission report treats it as a lander subsystem (GNC - TRN).; 2018-08-08 |
| Timeline | first funding 2018-08 (Tipping Point selection); the funding of the FO-flown 2014 predecessor is not in public sources; first flight test 2014 (predecessor autolanding system on Xombie, FO); 2022 (OPAL on a King Air aircraft); selected 2018-08-08 (Tipping Point selection, with the Peregrine demonstration as its goal; contract kickoff May 2019); launch 2024-01-08; landing not in public sources |
| On the Moon | partly operated |
| FO | tested |
Flights before the Moon¶
- FO: Masten Xombie rocket-powered lander, Mojave: Astrobotic Autolanding System (AAS), FO technology T0067 (TechPort 93996), 2014 (flights on 21 Feb, 12 Jun and 20 Jun 2014). Astrobotic's earlier autolanding system, which used terrain relative navigation and hazard detection to steer a rocket-powered lander to a safe spot. FO says the tests let Astrobotic begin a space-rated version for its lunar lander. Not shown to be OPAL's direct predecessor: OPAL is a new sensor that also draws on JPL's Mars 2020 work.
"Astrobotic Technology successfully demonstrates Terrain Relative Navigation and on-line hazard detection and identification/selection of safe landing locations in a flight-relevant environment"
— FO annual report, FY2014, p. 77, highlighted resultData: date: June 20, 2014
— FO annual report, FY2014, flights of TechPort project 93996 (pp. 7, 48)"These successful tests made possible by Flight Opportunities has enabled Astrobotic to begin work on a space-rated version of the AAS system for their commercial lunar lander."
— FO annual report, FY2014, p. 52
- aircraft (not FO): King Air B200 aircraft over the northern Mojave Desert, up to 9 km altitude, 2022 (announced 3 Nov 2022). The OPAL hardware and flight software itself, on a week-long campaign along more than 100 km of flight path that mimicked Peregrine's powered descent. It produced valid real-time position estimates; lessons went into the Peregrine unit.
"was fully validated during a week-long terrestrial flight test campaign above the mountains of the northern Mojave Desert in California"
— Landing tech: terrain relative navigation validated, ready for spaceflight (Astrobotic)"Throughout this test, OPAL flew aboard a King Air B200 aircraft up to 9km in altitude."
— Landing tech: terrain relative navigation validated, ready for spaceflight (Astrobotic)"Lessons learned from this flight test are being implemented into the OPAL system to improve performance for operation during the Astrobotic PM1 mission."
— Landing tech: terrain relative navigation validated, ready for spaceflight (Astrobotic)
Heritage hardware¶
- design heritage (Astrobotic and JPL TRN, Mars 2020): OPAL builds on Astrobotic's and JPL's earlier TRN systems. JPL transferred Mars 2020 Lander Vision System know-how and provided flight-proven FPGA computer-vision modules; Moog supplied avionics and structure based on its flight-proven hardware.
"Builds on Astrobotic's and JPL's prior TRN systems"
— NTRS 20230013300, page 3"Transfer Mars 2020 LVS + Feature Descriptor IP Implementation to Industry"
— NTRS 20230013300, page 5"JPL provided flight-proven FPGA computer vision modules for TRN and architectural and testing guidance."
— TechPort project 116335, outcomes, closeout
Funding before the lunar flight¶
- NASA STMD Tipping Point contract to Astrobotic for a TRL 9 TRN and visual velocimetry sensor, demonstrated on a lunar lander (80LARC19C0008, via NASA Langley; TechPort 116335, GCD). Lifetime obligation shown. Wikipedia calls it a $10 million Tipping Point; Tipping Points are cost-shared, and Astrobotic reported an overrun covered with other funds. (STMD Tipping Point (Game Changing Development); 2019-05 to 2024-07): $7,973,959
"WORK DELIVERS A TRL 9 TERRAIN RELATIVE NAVIGATION AND VISUAL VELOCIMETRY SENSOR TRN SENSOR PRODUCT BY DEMONSTRATING THE TRN SENSOR OPERATION ON A LUNAR LANDER MISSION."
— USAspending.gov, award 80LARC19C0008, description; obligated 7973959.0"Astrobotic expects an overrun (due to launch delays, landing site change, high hardware costs, and COVID). No work stoppage – other funds used to cover overrun."
— NTRS 20230013300, page 14"The TRN sensor was being developed under a $10 [[1,000,000|million]] NASA Tipping Point contract"
— Wikipedia article "Peregrine Mission One", revision 1368288193 of 2026-08-08T04:52:48Z
- Related, not the Peregrine unit: Astrobotic NASA SBIR 'Ultra-Compact Standalone Visual Relative Navigation' (194200 / 80NSSC20C0175, Phase II, 2020; Phase I 194200 / 80NSSC19C0288, 2019, $123,074). Linked by Astrobotic's 2023 review, which says a Phase II SBIR (Goddard) supports porting its TRN software to lower size, weight and power hardware: a next generation. (NASA SBIR (to Astrobotic); 2020-2022): $749,997
"Ultra-Compact Standalone Visual Relative Navigation"
— SBIR.gov award data, Phase II, 194200 / 80NSSC20C0175, award amount 749997.0"Internal R&D program supported by Phase II SBIR (GSFC) to port Astrobotic's TRN software to"
— NTRS 20230013300, page 5
How it got onto CLPS¶
"The system has been developed under a NASA Tipping Point contract that kicked off in May 2019."
— TechPort project 116335, outcomes, closeout (GCD program)"The OPAL Sensor shall be integrated with the Peregrine lander"
— NTRS 20230013300, page 2, Objective #1"Astrobotic was to demonstrate its standalone Terrain Relative Navigation (TRN) sensor as a payload on its first mission to the Moon."
— Wikipedia article "Peregrine Mission One", revision 1368288193 of 2026-08-08T04:52:48Z, Wikipedia, revision of 8 Aug 2026"Astrobotic Technology, Inc., Pittsburgh, Pennsylvania, $10 million"
— NASA Announces New Partnerships to Develop Space Exploration Technologies - NASA, release dated Aug 08, 2018"This project will culminate in a lunar technology demonstration mission, advancing a low-cost, reliable, high-performance, stand-alone Terrain Relative Navigation (TRN) sensor suite."
— NASA Announces New Partnerships to Develop Space Exploration Technologies - NASA"Selections are based on the agency’s third competitive"
— NASA Announces New Partnerships to Develop Space Exploration Technologies - NASA"The Astrobotic team will test the sensor as a technology demonstration on the company’s commercial lander, Peregrine."
— Astrobotic Selected for NASA Award to Develop Sensor for Precise Planetary Landings"The TRN sensor was being developed under a $10 [[1,000,000|million]] NASA Tipping Point contract with NASA [[Johnson Space Center]], [[Jet Propulsion Laboratory]], and Moog."
— Wikipedia article "Peregrine Mission One", revision 1368288193 of 2026-08-08T04:52:48Z
On the Moon¶
Peregrine could not land, so OPAL never did its job in descent. Astrobotic rates the TRN subsystem TRL 7 after the mission because it was not exercised for descent and landing in a lunar environment, while the rest of the lander's guidance and navigation reached TRL 9. NASA's TechPort record lists TRL 9 and calls the flight a technology demonstration in space; the two disagree. The transit checkouts planned for OPAL (star alignment, lunar exposure) are not reported in public sources.
"Not exercised for descent and landing in a lunar environment"
— Peregrine: Mission 1, August 2024 Post-Mission Report (Astrobotic), Figure 9, 'GNC – TRN', TRL 7"culminated in a technology demonstration of the sensor on Astrobotic's first mission in January 2024"
— TechPort project 116335, outcomes, closeout (trlCurrent 9)
FO's role¶
Grade: tested. FO flew Astrobotic's earlier autolanding system, which demonstrated terrain relative navigation on a rocket-powered lander in three Xombie flights in 2014. NASA says: 'Early testing of the system in 2014 under NASA’s Flight Opportunities program on a Masten Space Systems vehicle provided the foundation for the fundamental algorithms and system architecture.' FO's FY2014 report says the tests let Astrobotic begin 'a space-rated version of the AAS system for their commercial lunar lander'. So the FO-flown system is OPAL's direct predecessor in algorithms and architecture. The hardware is new (Moog avionics, JPL vision modules), and OPAL's own flight test (2022) was on a King Air aircraft, not an FO flight. The first pass graded this related, because it had no source that called the 2014 system OPAL's direct predecessor; an independent second check found NASA's statement and raised the grade to tested. Fit is retroactive for the 2014 flights: they predate CLPS.
FO flights: not in public sources. Three Xombie flights in 2014. The FY2016 report's line on testing in 2015 does not match FO's FY2014 and FY2015 campaign tables, and the follow-on 'Flight Two' project was canceled without flying (second check). Their cost is in FO records.
"Autolanding for Robotic Precursor Missions"
— TechPort project 93996, title (FO program, Astrobotic, 2013-2016)"These successful tests made possible by Flight Opportunities has enabled Astrobotic to begin work on a space-rated version of the AAS system for their commercial lunar lander."
— FO annual report, FY2014, p. 52"Builds on Astrobotic's and JPL's prior TRN systems"
— NTRS 20230013300, page 3"Astrobotic proposes to mature its Autolanding System (the AAS) from TRL 4 to 6 and evaluates performance through a series of tests on an sRLV. The AAS is a complete system for GPS-denied navigation"
— TechPort project 93996, description (FO project 93996, Astrobotic Technology, 2013-2016, Xombie flights 2014)"Astrobotic Technology successfully demonstrates Terrain Relative Navigation and on-line hazard detection and identification/selection of safe landing locations in a flight-relevant environment using Masten Space Systems’ Xombie vehicle."
— FO annual report, FY2014, technologies, T0067, outcomes (annual report p. 77)"These successful tests made possible by Flight Opportunities has enabled Astrobotic to begin work on a space-rated version of the AAS system for their commercial lunar lander."
— FO annual report, FY2014, entry 93996, outcomes, p. 52"Early testing of the system in 2014 under NASA’s Flight Opportunities program on a Masten Space Systems vehicle provided the foundation for the fundamental algorithms and system architecture."
— NASA, Astrobotic Advance Precision Landing Systems in Plane Test - NASA, NASA JSC, Sep 01, 2022, about Astrobotic's TRN system for Peregrine"Astrobotic has been developing its commercial TRN system for over a decade."
— NASA, Astrobotic Advance Precision Landing Systems in Plane Test - NASA
Who built it¶
"said Dr. Andrew Horchler, Chief Research Scientist at Astrobotic and principal investigator for OPAL"
— Landing tech: terrain relative navigation validated, ready for spaceflight (Astrobotic), Astrobotic, 3 Nov 2022"Astrobotic is the prime for this work which included subcontracts for Moog, Inc., JPL, and NASA JSC."
— TechPort project 116335, outcomes, closeout
Moon Base need¶
"technologies should enable safe and accurate landings in all visibility conditions"
— Moon to Mars knowledge base: ESDMD #1101: Lunar Precision Landing and Hazard Avoidance for Human Exploration, Description"The goal is a final touchdown accuracy of 100 m during future closed-loop lunar landings"
— NTRS 20230013300, page 2 (FY23 GCD annual program review)
Open questions¶
- Which parts of the FO-flown Autolanding System carry into OPAL? NASA says the 2014 FO tests 'provided the foundation for the fundamental algorithms and system architecture'; the hardware is new.
- What did OPAL do in transit (star alignment and lunar exposure checkouts)? Not in public sources.
- TRL after Peregrine: Astrobotic says 7, TechPort says 9. Which is NASA's view?
- The Tipping Point's total value and Astrobotic's cost share: Wikipedia says $10 million; USAspending shows $7,973,959 obligated by NASA. The split is not in public sources.
- Did FO fly Astrobotic's landing system in 2015? One line of the FY2016 report says so; FO's campaign tables show only the 2014 flights (second check).