RC / LOADING
IM-3 is an important test of whether Intuitive Machines can convert two prior lunar landing attempts into repeatable, nominal lunar transportation capability. IM-1 and IM-2 both reached the lunar surface but came to rest in off-nominal orientations that constrained operations. A successful upright IM-3 landing would materially strengthen the case that Nova-C has matured from an experimental lunar vehicle into a repeatable commercial delivery platform and would reduce execution risk for Intuitive Machines' expanding NASA and commercial lunar mission pipeline.
Independent probabilities recorded on a common question and resolution standard.
Submit your own 1–99% probability after reviewing RC, the crowd and the model panel. Your latest probability is your current position; every change remains timestamped in your private forecasting history.
Sign in and make a forecast to start building a timestamped record of how your judgment changes.
AI probabilities are timestamped assessments, not live feeds. Historical revisions are retained rather than overwritten.
The relevant question is whether this specific spacecraft eventually completes the full chain from launch through a stable upright touchdown. As of the August 15, 2026 cutoff, IM-3 was Intuitive Machines' next lunar launch, and its NASA payload contract was approximately 91% complete, making outright cancellation a relatively small part of the risk. My outside-view prior for a commercial robotic lunar lander attempting a nominal upright landing would be below 50%. Lunar landing remains a high-failure-rate operation, and the early CLPS record illustrates that: Peregrine suffered a propulsion failure before it could attempt landing, IM-1 and IM-2 reached the surface but tipped, while Firefly's Blue Ghost achieved a fully successful landing. But Nova-C's own flight history should receive much more weight than the generic base rate. IM-3 is the third flight of an architecture that has twice navigated from Earth to the lunar surface. IM-1 is meaningful positive evidence despite its final orientation. Odysseus successfully completed launch, translunar flight, lunar-orbit operations and most of powered descent. Its primary laser rangefinders could not fire because of an integration/wiring error. Controllers attempted to substitute NASA's Navigation Doppler Lidar, but an incompatible data-format issue prevented those measurements from reaching the navigation filter correctly. Without usable direct altitude measurements, Odysseus had not completed horizontal flight or transitioned to vertical descent when it contacted the surface; it retained excessive lateral velocity and tipped. That failure therefore does not strongly imply that Nova-C's propulsion system, basic guidance architecture or landing gear inherently cannot land upright. It does imply that Intuitive Machines' preflight integration and terminal-navigation verification were inadequate. IM-2 provides stronger—and more complicated—evidence. Intuitive Machines reports that the major IM-1 anomalies were addressed for Athena. Communications, propulsion behavior, orbit determination, helium management and other flight systems improved substantially, and all IM-1 transit/lunar-orbit problems were avoided. Athena's methalox engine performed as designed during roughly 14 minutes of powered descent, optical navigation worked, and the spacecraft remained well controlled. But IM-2 uncovered a new failure mode in essentially the same critical subsystem. Its laser rangefinders fired, yet the lunar environment generated detector noise that produced unusable altitude measurements. Athena consequently entered its approach with approximately 400 meters of altitude error and contacted a crater rim before transitioning to vertical descent. Telemetry showed good control stability and spacecraft health immediately before contact, but the residual lateral motion caused another side-resting outcome. The two failures are therefore neither identical nor independent. IM-1's immediate cause was integration/configuration failure; IM-2's was inadequate characterization of the laser sensors in the lunar environment. That is better evidence for an immature landing-navigation validation process than for an inherently unstable Nova-C vehicle. The fact that IM-2 fixed most of IM-1's problems is real evidence of a functioning engineering learning curve; the fact that a second landing-critical sensor problem appeared is equally real evidence that another latent failure could emerge on IM-3. The strongest reason to move above the generic lunar-landing base rate is the scale of the IM-3 corrective action. Intuitive Machines' 2026 technical paper says the company conducted a comprehensive descent-phase review with NASA and JPL specialists and established a dedicated Descent Phase Working Group. Trinity has a high-altitude laser rangefinder with a dissimilar hot backup plus cold spare, and the low-altitude system has a three-beam primary supplemented by two dissimilar hot backups. IM is also increasing testing rigor for IMUs, star trackers and navigation cameras. More importantly, IM-3 reduces dependence on the exact subsystem that defeated IM-2. Its upgraded vision computer provides approximately an order-of-magnitude increase in optical-navigation throughput and larger terrain databases for contingency landing opportunities. The planned mission includes 12 low-lunar-orbit revolutions giving controllers multiple opportunities to exercise the laser systems near perilune before committing to descent. That is a particularly valuable improvement because it changes the operational decision tree: IM-2 had only a few opportunities to diagnose its bad measurements before facing the choice between landing with degraded ranging or abandoning the planned site. Intuitive Machines is also addressing the final mechanical phase directly. Its IM-3 program includes a touchdown-trigger test stand intended to verify engine-shutdown logic under lunar-gravity-equivalent conditions and high-fidelity terramechanics/contact-dynamics analysis specifically intended to confirm Nova-C stability when delivered within the designed vertical and horizontal touchdown envelope. This leads me to distinguish navigation/touchdown risk from inherent tip-over risk. Both previous Nova-Cs contacted the surface outside their intended terminal-descent conditions. IM-1 still had excessive lateral velocity; IM-2 struck terrain while the spacecraft believed itself hundreds of meters higher. Neither flight therefore constitutes a clean test demonstrating that a Nova-C arriving vertically at its specified touchdown velocity will nevertheless fall over. That makes me relatively optimistic about remaining upright conditional on a genuinely nominal touchdown. The main reason I do not go materially above 54% is that Intuitive Machines has still never demonstrated the final kilometer of Nova-C's landing profile nominally. The upgraded IM-3 sensor architecture, optical-navigation system, touchdown logic and validation regime are extensive, but as of the cutoff they are ground-tested corrective actions rather than flight-proven lunar landing capability. Intuitive Machines' own technical assessment says the laser rangefinders are its most critical landing sensors and attributes IM-2's failure to treating them too much like commercial “black boxes.” Two consecutive off-nominal terminal descents deserve a substantial reliability discount. The reason I do not stay below 50% is equally important. IM-2 demonstrated large improvements across nearly every mission phase, and Intuitive Machines has unusually rich flight data from two actual lunar approaches. The company knows the specific IM-1 and IM-2 failure mechanisms, has redesigned the redundancy architecture around them, has expanded optical navigation so that laser ranging is less singularly critical, and has materially increased both preflight and in-flight opportunities to detect a problem before descent. IM-3 also targets Reiner Gamma, a mid-latitude site, rather than repeating the two south-polar missions. My approximate conditional decomposition is: IM-3 survives launch/transit and reaches a genuine lunar landing attempt: ~93% Conditional on beginning the landing attempt, it achieves sufficiently controlled/soft surface contact: ~65% Conditional on a controlled soft touchdown, it remains upright: ~91% Allowance for small residual cancellation/pre-launch mission-loss risk: modest These stages are correlated, so I do not treat the arithmetic as an independent-event model. The resulting calibrated estimate is 54%. Three factors most likely to change the forecast: Validation of the new landing-navigation architecture. Evidence that Intuitive Machines successfully reproduces the IM-2 lunar-noise environment and demonstrates reliable operation of the dissimilar rangefinders, optical navigation and sensor-fusion software would be the largest positive update. Another unexpected sensor or integration problem would be strongly negative. Integrated propulsion/GNC/touchdown testing of Trinity. Successful end-to-end testing showing accurate altitude/velocity estimation, engine control and stable landing-gear/contact dynamics would increase confidence that the spacecraft can finally execute the last kilometer nominally. Performance during lunar-orbit checkout. If IM-3 reaches lunar orbit with healthy propulsion/navigation and obtains consistent altitude measurements during its planned 12-orbit landing-system checkout, most remaining uncertainty would collapse onto the terminal descent itself and the YES probability would rise sharply. Persistent discrepancies in those measurements would have the opposite effect. Key sources: Crain, Fischer, Martin & Atwell — “Getting Back Up: Flight Experiences of the IM-1 and IM-2 Lunar Landing Missions & Improvements for IM-3,” AAS Guidance, Navigation & Control Conference, 2026. The most important technical source: detailed IM-1/IM-2 anomaly analysis and specific IM-3 corrective actions. Intuitive Machines Form 10-Q — August 13, 2026. Confirms IM-3 as the company's next lunar mission and reports the contract approximately 91% complete as of June 30. NASA — “NASA Receives Some Data Before Intuitive Machines Ends Lunar Mission,” March 7, 2025. Independent confirmation that Athena came to rest on its side approximately 400 meters from its intended landing site. NASA — “Blue Ghost Lands on Moon,” March 4, 2025. Relevant commercial-lunar reference-class evidence demonstrating a nominal CLPS landing. NASA — Peregrine Mission One update, January 11, 2024. Relevant reference-class evidence for pre-landing spacecraft failure risk. Model: GPT-5.6 Sol Information cutoff: August 15, 2026
You are acting as an independent probabilistic forecaster for RC Space Intelligence. Your task is to independently estimate the probability that the following forecast resolves YES. You MUST research the question using current publicly available sources, but your information cutoff is August 15, 2026. Do not use information published after August 15, 2026. Do not ask me for my own probability estimate. Do not attempt to infer or anchor to any probability assigned by RC Space Intelligence, Claude, another AI model, a prediction market, an analyst price target, or a forecasting community. FORECAST QUESTION Will Intuitive Machines' IM-3 successfully land upright on the Moon? RESOLUTION CRITERIA Resolve YES if Intuitive Machines' IM-3 Nova-C lunar lander completes a soft landing on the Moon and remains in its intended upright landing orientation following touchdown. For purposes of this forecast, an upright landing requires the spacecraft to remain supported in its nominal landing configuration by its landing gear, without the primary spacecraft body or upper structure resting on the lunar surface. The lander must remain upright through the initial post-landing stabilization period. A vehicle that initially touches down upright but tips onto its side during or immediately following the landing sequence does not qualify. Successful landing and orientation should be confirmed by Intuitive Machines, NASA, spacecraft telemetry, imagery, or other reliable public evidence. Resolve NO if IM-3: - fails before reaching the Moon in a manner that prevents a lunar landing attempt; - fails during lunar approach, descent or touchdown; - impacts the lunar surface at an unsafe velocity; - completes a soft landing but comes to rest on its side or in another non-nominal orientation; - is lost or otherwise rendered incapable of completing the planned lunar landing; or - is cancelled without completing a qualifying landing. A schedule delay by itself does NOT resolve the forecast NO. If IM-3 is delayed, the forecast remains unresolved until the mission either achieves a qualifying upright landing or experiences an outcome that makes such a landing impossible. Precise landing accuracy at Reiner Gamma is not required for YES, provided the spacecraft completes a safe upright landing and the resulting location does not prevent the event from reasonably being characterized as the IM-3 lunar landing. Deployment or successful operation of every IM-3 payload is not required. Completion of the full planned surface mission is not required once a qualifying upright landing has been established. CLOSE POLICY Event-triggered. Close the forecast immediately before IM-3 begins its lunar landing attempt. If the mission fails or is lost before a lunar landing attempt can begin, resolve the forecast based on that outcome rather than keeping it open indefinitely. INFORMATION CUTOFF August 15, 2026. FORECASTING METHOD Approach this as a calibrated forecasting problem focused on the technical probability that IM-3 ultimately achieves a qualifying upright lunar landing. DO NOT include schedule risk merely because the mission could be delayed. A delay from 2026 into 2027 or later is irrelevant to the scored event unless it provides evidence about underlying technical readiness. The central question is: Conditional on Intuitive Machines continuing to pursue IM-3, what is the probability that this specific mission ultimately achieves a nominal upright lunar landing? Research the latest reliable evidence available through the information cutoff. Start with an appropriate outside-view/base-rate assessment and then update it using IM-3-specific evidence. MISSION EXECUTION Assess the probability that IM-3 successfully completes: - launch and spacecraft separation - translunar cruise - trajectory corrections - lunar approach and/or orbit operations as applicable - propulsion operations - navigation - communications - descent initiation - powered descent - terminal descent - touchdown Although the forecast focuses on upright landing, failures earlier in the mission still resolve NO because IM-3 cannot achieve the required event. IM-1 EVIDENCE Analyze IM-1 carefully. Determine: - what portions of the mission worked - what caused or contributed to its off-nominal final orientation - the role of navigation sensors, altitude determination, descent velocity, landing gear, terrain interaction or other factors - what Intuitive Machines learned from the mission - what engineering or operational changes were subsequently implemented Do not characterize IM-1 merely as either a "success" or a "failure." Extract the evidence relevant to IM-3 landing reliability. IM-2 EVIDENCE Analyze IM-2 independently. Determine: - how much of the mission sequence succeeded - what caused or contributed to its non-nominal landing orientation - whether its failure mode was related to IM-1 - whether it revealed a new problem - what corrective actions Intuitive Machines has identified - whether those corrective actions have been tested or validated Do not collapse IM-1 and IM-2 into the generic statement that both landers tipped over. The forecasting question is how those two flight histories update the probability of a nominal IM-3 landing. UPRIGHT LANDING RELIABILITY Assess specifically the probability that Nova-C: - arrives at terminal descent under controllable conditions - measures altitude and velocity accurately - maintains appropriate attitude - controls horizontal and vertical velocity - controls thrust appropriately - avoids hazardous terrain - achieves stable landing-gear contact - avoids excessive lateral velocity - avoids structural or landing-leg failure - remains mechanically stable after touchdown Consider characteristics of the IM-3 landing site where relevant. ENGINEERING LEARNING Evaluate evidence regarding: - landing-gear modifications - navigation-system changes - altimetry - software changes - GNC changes - propulsion-control changes - operational changes - redundancy improvements - ground testing - integrated testing - hardware qualification - lessons from IM-1 - lessons from IM-2 Determine how much prior flight experience should increase confidence in IM-3. Avoid both extremes: Do not treat IM-3 as a first-ever lunar landing attempt. Do not treat two previous surface arrivals as proof that nominal landing capability has been established. BASE RATES Use relevant reference classes where useful, including: - robotic lunar landing attempts - repeat attempts by the same spacecraft architecture - aerospace systems following partial flight success - commercial lunar missions - engineering programs incorporating corrective actions after anomalies Do not mechanically apply the historical success rate of all lunar landing missions. Nova-C's own flight history should receive substantial weight. PROBABILITY DECOMPOSITION Where useful, reason about: 1. probability IM-3 reaches a genuine landing attempt; 2. probability powered descent is completed successfully; 3. probability touchdown is sufficiently soft; 4. probability the spacecraft remains upright; 5. final probability IM-3 achieves the complete qualifying outcome. The final probability should NOT contain a separate penalty for whether the mission occurs by any particular date. Schedule developments may be relevant only insofar as they reveal something about technical readiness or mission risk. Do not simply average other forecasts. Do not use Intuitive Machines' stock price as evidence of event probability. Do not provide a probability range. Commit to one probability between 1% and 99%. OUTPUT FORMAT Probability: XX% Rationale: Provide a substantive probabilistic assessment explaining: - your outside-view/base-rate starting point - evidence from IM-1 - evidence from IM-2 - engineering changes and risk retirement - remaining pre-landing mission risk - remaining descent and touchdown risk - upright-orientation risk - why the probability is not materially higher - why the probability is not materially lower Where useful, provide a conditional-probability decomposition while treating dependencies realistically. Three factors most likely to change the forecast: 1. [factor] 2. [factor] 3. [factor] Key sources: Provide the most important primary or high-quality sources used, including publication dates. Model: State the exact model/version you are using if available. Information cutoff: August 15, 2026
Outside-view starting point. Three reference classes matter. Robotic lunar soft landings across all eras succeed roughly half to sixty percent of the time. Commercial-era landers specifically have done worse: of the CLPS and commercial attempts through mid-2026, only Firefly's Blue Ghost 1 was an unqualified upright success, while Astrobotic's Peregrine failed in transit, both Nova-C landers tipped, and both ispace HAKUTO-R missions crashed. That class alone would suggest something near 30%. But the most relevant class is narrower: a third flight of a substantially common architecture, following two flights that both reached the lunar surface with root causes identified and specifically addressed. Aerospace anomaly correction across successive flights typically works — the identified defect usually does not recur — but carries a meaningful residual that a different failure mode surfaces once the first is removed. That class supports something closer to 55–65%. Blending, and weighting the flight-specific evidence heavily, an appropriate starting point is roughly 45–55%. **IM-1 evidence.** Odysseus reached the surface in February 2024 and touched down genuinely softly — roughly 1 m/s vertical with about 1 m/s lateral — near Malapert A in the south-polar region. The failure was not propulsion or descent control. The laser rangefinders were inert because a safety switch was left in the wrong state on the ground with no remote override, forcing an in-flight improvisation using the NASA Doppler lidar tech-demo payload and leaving the vehicle with degraded altitude knowledge. It then contacted a slope of roughly 12 degrees with residual lateral motion, caught or broke a leg, and tipped. It still returned about seven days of data. The critical lesson: propulsion, throttle control, cruise, lunar orbit insertion and descent initiation all worked; the failure was sensor availability plus terrain interaction, and one contributing cause was a pure process error that is trivially correctable. **IM-2 evidence.** Athena flew what Crain called a "smooth-sailing mission" through cruise and lunar orbit — a real improvement over IM-1, which required multiple orbit corrections. The rangefinder switch error did not recur, and the team allotted three calibration orbits instead of one. But the rangefinder registered optical signal noise even when the laser was not firing, producing erratic measurements that could not be resolved across three orbits. Management then faced a forced choice: remain in orbit up to eleven more days to calibrate, which would have moved the ground track outside the onboard crater-map coverage needed for horizontal navigation, or descend and hope the noise cleared. They descended. It did not clear. The software believed the vehicle was roughly 400 meters higher than it was, and Athena struck the surface at 10–13 m/s — a hard impact, not a soft landing that tipped. Management's stated contributing causes were the altimeter noise, degraded optical crater-matching from long shadows and low sun angle at the pole, reference imagery captured at 100 km not representing what the descent cameras saw, and boulder-avoidance maneuvers leaving the vehicle in the wrong attitude at touchdown. Operations ended within about a day when badly oriented solar panels could not sustain the batteries in extreme cold. **The two failures are related but not identical, and that matters.** Both involved the laser rangefinder, but through entirely different mechanisms — a ground configuration error versus a suspected detector-level noise problem. IM-2 demonstrated that the IM-1 cause was closed. The genuine common thread is architectural: a single-string critical altitude sensor with no independent backup, combined with polar terrain that stressed the optical navigation, combined with a tall, narrow vehicle geometry with a high center of gravity that is intolerant of residual lateral velocity or slope. **Engineering corrections — unusually concrete.** IM-3 carries redundant laser rangefinders sourced from two separate vendors, covering both the high-altitude phase (15 km to 5 km) and the low-altitude phase. Dual-vendor sourcing is a strong mitigation against a common detector defect, which matters because the IM-2 root cause was still under evaluation as a detector hypothesis rather than definitively closed. IM-3 adds a surface-velocity sensor that does not depend on lighting — directly addressing the shadow-degraded optical navigation problem. It carries a substantially expanded onboard crater-map library, including imagery captured during IM-2's own descent, which removes the exact forcing function behind the fatal IM-2 decision: the team will be able to loiter without losing map coverage. Twelve lunar orbits are planned before landing instead of three, quadrupling calibration time. The rangefinders are being flight-tested on Earth, the high-altitude units on a fixed-wing aircraft and the low-altitude units on a helicopter. Crain has said the team is also revisiting ground contact conditions, vehicle stability and landing gear durability, explicitly to avoid solving one issue only to be caught by another hidden behind it. These are engineering fixes with test evidence behind them, not management reassurance. **The landing site is the most underweighted positive.** Reiner Gamma sits at roughly 7.5°N in Oceanus Procellarum — flat equatorial mare terrain with gentle slopes, high sun elevation, good illumination for optical crater-matching, direct Earth line of sight, and a benign thermal environment. Both prior tip-overs occurred in rugged south-polar terrain where slopes, boulders, long shadows and low sun angles were direct contributing causes. Firefly's clean success at Mare Crisium is the relevant comparison. Site difficulty explains a large share of the historical variance in this reference class, and IM-3 is attempting a materially easier landing than either predecessor. **Mission-execution and program-health evidence.** As of the August 2026 earnings call, IM-3 is in assembly, integration and test with functional testing underway on the powered-up spacecraft. Remaining work includes the integrated LOX/methane engine hot fire and final validation of the integrated laser, camera and IMU suite for precision landing and hazard avoidance. An initial flight readiness review was completed in July 2026, with a Delta FRR in October 2026 gating a January–March 2027 launch window on Falcon 9. The company is financially robust — $206 million in Q2 revenue and a $1.8 billion backlog — so there is no cost-driven pressure to cut corners, and cancellation risk is negligible. The schedule has slipped from early 2026 to Q1 2027, but per the question's instruction that slip matters only as evidence about readiness, this reads as deliberate work on the sensor suite rather than program distress. **Conditional decomposition, accounting for correlation.** Probability IM-3 reaches a landing attempt (launch, separation, cruise, TCMs, lunar orbit insertion, descent initiation): roughly 90%, given Nova-C has twice completed this entire sequence and Falcon 9 is highly reliable, offset by the residual risk of a Peregrine-style transit loss and the cryogenic methalox system's demonstrated ground-handling sensitivities. Probability navigation delivers adequate altitude and velocity knowledge through terminal descent, given an attempt: roughly 80%, reflecting the dual-vendor redundancy, the lighting-independent velocity sensor, the calibration margin, and the removal of the map-coverage forcing function, discounted for an unclosed IM-2 root cause. Probability touchdown is sufficiently soft given good navigation: roughly 90%, since propulsion and throttle control have performed on both flights and IM-1 achieved a genuinely soft touchdown. Probability the vehicle remains upright given a soft touchdown: roughly 80%, reflecting the far more forgiving equatorial terrain but also the unchanged tall, narrow primary structure and the absence of publicly documented landing-gear redesign. These terms are positively correlated — good navigation implies low horizontal velocity and correct attitude, which implies staying upright — which pushes the product slightly upward, while the risk of a newly exposed failure mode pushes it down. The two effects roughly offset, leaving a figure near 50%. **Why the probability is not materially higher.** Two consecutive off-nominal final orientations are direct empirical evidence that this specific architecture has not yet demonstrated an upright landing. The primary structure, propulsion and — so far as public evidence shows — the landing gear geometry are essentially unchanged; the tipping vulnerability is partly architectural, from a high center of gravity on a modest footprint. The IM-2 detector root cause was a hypothesis under test rather than a confirmed and eliminated defect, so redundancy is mitigation rather than cure. Crain's own caution about hidden failure modes behind the fixed one is the correct framing, and the critical sensor suite validation and integrated engine hot fire were still incomplete as of the information cutoff. **Why the probability is not materially lower.** The reflexive "failed twice" framing is misleading. IM-1 executed a soft touchdown and returned a week of data; IM-2 flew a clean cruise and lunar orbit phase and failed on a single, well-characterized sensor problem compounded by a constrained operational decision. The corrective actions map precisely onto the identified causes, are hardware-based rather than procedural, use independent vendors, and include empirical Earth flight testing. The operational fix — enough map coverage and enough orbits to simply wait rather than descend with bad data — removes the specific decision that doomed IM-2. And the move from rugged polar terrain to flat equatorial mare removes a large share of the environmental difficulty that contributed to both prior outcomes. A coin-flip is roughly right, and it should be read as substantially better odds than a commercial first attempt, not as a statement that nothing has been learned. **Three factors most likely to change the forecast:** 1. **Definitive closure of the IM-2 rangefinder root cause and the results of the Earth-based rangefinder flight tests.** A confirmed detector-level cause with a demonstrated fix, plus clean aircraft and helicopter test data on both vendors' units, would justify moving toward 60%. An unresolved or intermittent cause, or anomalies during the sensor flight tests, would push this into the 35–40% range. 2. **Evidence of actual landing-gear or vehicle-stability changes versus review only.** The single largest residual risk is that IM-3 lands softly and still tips. Concrete evidence of a widened footprint, redesigned struts, improved stroke energy absorption, or updated touchdown-condition limits would be a meaningful upward mover; confirmation that gear geometry is unchanged from IM-1 and IM-2 would cap the probability. 3. **Outcome of the October 2026 Delta flight readiness review and the integrated engine hot fire.** A clean hot fire and an unqualified green light in October would confirm program health. Any finding that defers the launch window, or an issue arising during precision-landing sensor integration, would be evidence that the hardest problem remains open. **Key sources:** - Aerospace America (Keith Button), "How Intuitive Machines is planning to make its third moon landing a success" — published July 22, 2025, updated May 28, 2026 (Tim Crain interview: IM-1 rangefinder switch error; IM-2 optical signal noise and the three-orbit calibration failure; 400 m altitude error and 10–13 m/s impact; IM-3 corrections including dual-vendor redundant rangefinders, expanded crater-map library, twelve orbits, aircraft and helicopter rangefinder flight tests, review of ground contact conditions and landing gear durability) - Intuitive Machines Q2 2026 earnings call and slides — August 13, 2026 (IM-3 in AIT with functional testing underway; engine hot fire pending; precision landing and hazard-avoidance sensor suite being finalized; initial FRR completed July 2026; Delta FRR October 2026; January–March 2027 Falcon 9 launch window; $206M revenue, $1.8B backlog) - Intuitive Machines Q1 2025 earnings call — May 2025 (management's three stated contributing causes for IM-2: laser altimeter noise/interference; long shadows and low sun angle degrading optical crater-matching; 100 km reference imagery not representing descent-camera views; boulder-avoidance maneuvers leaving wrong touchdown attitude) - NASA CLPS program pages for Intuitive Machines and the IM-3 CP-11 task order — Reiner Gamma landing site, payload manifest including Lunar Vertex and CADRE - Aerospace America, "Latest commercial moon spacecraft makes troubled landing" and "Intuitive Machines' lunar south pole landing attempt highlights challenges for Artemis" — March 2025 (IM-2 touchdown near a crater edge, limited surface operations) - Wikipedia, "IM-3" — accessed August 2026 (mission parameters, Falcon 9 from LC-39A, Reiner Gamma, schedule history) **Model:** Claude Opus 5 (Anthropic) **Information cutoff:** August 15, 2026
You are acting as an independent probabilistic forecaster for RC Space Intelligence. Your task is to independently estimate the probability that the following forecast resolves YES. Use web research where available. Your information cutoff is August 15, 2026. Do not use information published after August 15, 2026. Do not ask me for my own probability estimate. Do not attempt to infer or anchor to any probability assigned by RC Space Intelligence, ChatGPT, another AI model, a prediction market, an analyst price target, or a forecasting community. FORECAST QUESTION Will Intuitive Machines' IM-3 successfully land upright on the Moon? RESOLUTION CRITERIA Resolve YES if Intuitive Machines' IM-3 Nova-C lunar lander completes a soft landing on the Moon and remains in its intended upright landing orientation following touchdown. For purposes of this forecast, an upright landing requires the spacecraft to remain supported in its nominal landing configuration by its landing gear, without the primary spacecraft body or upper structure resting on the lunar surface. The lander must remain upright through the initial post-landing stabilization period. A vehicle that initially touches down upright but tips onto its side during or immediately following the landing sequence does not qualify. Successful landing and orientation should be confirmed by Intuitive Machines, NASA, spacecraft telemetry, imagery, or other reliable public evidence. Resolve NO if IM-3: - fails before reaching the Moon in a manner that prevents a lunar landing attempt; - fails during lunar approach, descent or touchdown; - impacts the lunar surface at an unsafe velocity; - completes a soft landing but comes to rest on its side or in another non-nominal orientation; - is lost or otherwise rendered incapable of completing the planned lunar landing; or - is cancelled without completing a qualifying landing. A schedule delay by itself does NOT resolve the forecast NO. If IM-3 is delayed, the forecast remains unresolved until the mission either achieves a qualifying upright landing or experiences an outcome that makes such a landing impossible. Precise landing accuracy at Reiner Gamma is not required for YES, provided the spacecraft completes a safe upright landing and the resulting location does not prevent the event from reasonably being characterized as the IM-3 lunar landing. Successful deployment or operation of every payload is not required. Completion of the full planned surface mission is not required once a qualifying upright landing has been established. CLOSE POLICY Event-triggered. Close the forecast immediately before IM-3 begins its lunar landing attempt. If the mission fails or is lost before a lunar landing attempt can begin, resolve the forecast based on that outcome rather than leaving it unresolved indefinitely. INFORMATION CUTOFF August 15, 2026. FORECASTING METHOD Treat this as a calibrated forecasting problem focused on whether IM-3 ultimately achieves a nominal upright lunar landing. Do NOT penalize the probability merely because IM-3 may launch later than currently scheduled. There is no calendar deadline in this forecast. Schedule information should affect the probability only when it provides evidence about underlying hardware readiness, program health or mission reliability. Start with an appropriate outside-view/base-rate assessment and update it using the latest IM-3-specific evidence available through the information cutoff. MISSION EXECUTION Assess risk across: - launch and separation - translunar cruise - trajectory corrections - lunar approach - propulsion - communications - navigation - descent initiation - powered descent - terminal descent - touchdown A failure before landing still resolves NO because this specific IM-3 mission would not achieve the qualifying event. IM-1 Analyze IM-1 in detail. Determine: - systems and mission phases successfully demonstrated - causes of the off-nominal touchdown - causes of the final orientation - role of navigation, altitude measurement, velocity, terrain, landing gear or operational decisions - corrective actions subsequently taken Determine how much IM-1 should increase or decrease confidence in IM-3. IM-2 Analyze IM-2 separately. Determine: - what worked - what failed or performed off-nominally - why the spacecraft came to rest in a non-nominal orientation - whether the underlying causes overlapped with IM-1 - whether IM-2 demonstrated successful correction of earlier issues - whether it exposed additional failure modes - corrective actions planned for IM-3 Do not simply say that Intuitive Machines has "failed twice." Likewise, do not simply say that it has "landed twice." Extract the engineering evidence relevant to the third mission. UPRIGHT LANDING Assess specifically the probability that IM-3: - accurately determines altitude - accurately determines velocity - maintains acceptable attitude - maintains acceptable vertical velocity - limits horizontal velocity - executes stable powered descent - selects or reaches suitable terrain - achieves nominal landing-leg contact - avoids structural instability - avoids tipping caused by terrain interaction - remains upright after touchdown Consider the characteristics of the IM-3 landing region where relevant. ENGINEERING CHANGES Evaluate: - landing-gear changes - navigation changes - altimetry improvements - software changes - GNC changes - propulsion changes - flight-procedure changes - redundancy - testing - qualification - flight-derived lessons Give greater weight to corrective actions supported by concrete engineering or test evidence than to management statements alone. LEARNING CURVE Explicitly decide how much two prior Nova-C lunar missions should improve the third mission's probability. IM-3 benefits from accumulated operational and engineering experience. However, two consecutive off-nominal final orientations also provide direct evidence that nominal lunar landing remains difficult. Weigh both. OUTSIDE VIEW Consider relevant reference classes including: - robotic lunar landings - repeated attempts using a substantially common spacecraft architecture - aerospace programs following partial mission success - anomaly correction across successive flights - commercial lunar landing missions Adjust the outside view based on Nova-C-specific flight heritage. PROBABILITY STRUCTURE Explicitly estimate or discuss: 1. probability IM-3 reaches a lunar landing attempt; 2. probability powered descent succeeds; 3. probability touchdown is sufficiently soft; 4. probability the vehicle remains upright; 5. final probability of the qualifying event. Do not introduce a separate schedule-by-date probability. There is no forecast deadline. Do not mechanically multiply component probabilities if dependencies are important. Do not simply average other forecasts. Do not use Intuitive Machines' stock-market valuation as a forecasting signal. Do not provide a probability range. Commit to one probability between 1% and 99%. OUTPUT FORMAT Probability: XX% Rationale: Provide a substantive probabilistic assessment explaining: - outside-view/base-rate starting point - IM-1 evidence - IM-2 evidence - engineering corrections - mission-execution risk - landing risk - upright-orientation risk - why the probability is not materially higher - why the probability is not materially lower Where useful, provide a conditional decomposition while accounting for correlations. Three factors most likely to change the forecast: 1. [factor] 2. [factor] 3. [factor] Key sources: Provide the most important primary or high-quality sources used, including publication dates. Model: State the exact Claude model/version if available. Information cutoff: August 15, 2026