1. The question left open
Entry 866 asked how many monitoring points a 1 km² sail needs. Entry 867 asks what the monitor is for. If the health monitor detects a puncture or a darkened patch, does the tug’s software use that information immediately to replan thrust? Or does it store the data for weeks, downlink it when Earth is in view, and wait for human operators to decide?
The answer depends on three clocks: the spacecraft’s control loop, the light-time to Earth, and the dynamics of the failure.
2. How fast does a sail control loop run?
Solar sails are slow spacecraft. The thrust is small, the inertia is large, and the flexible modes of the membrane are low-frequency. LightSail 2, a 32 m² LEO sail, used one reaction wheel and three magnetorquers for momentum management. It spent at most a few tens of minutes per orbit in a thrust-producing orientation. Its attitude control loop ran on timescales of seconds to minutes, but the orbit-changing effect accumulated over days to weeks.
NEA Scout, an 86 m² deep-space sail, planned a more deliberate rhythm. The guidance and control design documents describe trajectory correction maneuvers spaced by weeks or months, not hours. The spacecraft had reaction wheels, an active mass translator, and a cold-gas thruster system, but the dominant strategy was to hold a commanded attitude and let the sail produce a small, continuous thrust. Errors were corrected at scheduled waypoints.
Solar Cruiser, a much larger deep-space sail studied by NASA, faces even slower dynamics. Tyler (2023) notes that large sails have low-frequency flexible modes that hinder slewing and pointing. The bigger the sail, the longer it takes to change its attitude meaningfully. A 1 km² minimoon tug would likely have characteristic response times of hours to days for large attitude changes, and weeks to months for orbit-changing thrust arcs.
3. How fast does optical degradation matter?
Optical degradation is not a sudden event in most cases. Coating darkening, UV embrittlement, and hydrogen blistering accumulate over weeks and months. A puncture from a micrometeoroid is sudden, but the immediate effect — a local hot spot or a small thrust imbalance — may take days or weeks to grow into a mission-critical problem. A tear that propagates across a seam could be faster, but even a tearing polymer membrane propagates at finite speed under sail tension.
This slowness is good news for ground-in-the-loop management. If a reflectometer notices a 2% reflectivity drop over a month, there is no urgency measured in seconds. The data can be batch-downloaded, compared with the trajectory fit, and folded into the next long-range thrust plan.
4. Light-time changes the arithmetic
At 1 AU, round-trip light time is about 16 minutes. At 2 AU, it is 32 minutes. At Mars opposition, it can exceed 40 minutes. These are short compared with sail dynamics, so near-Earth and cislunar operations can remain ground-in-the-loop without much penalty. A human operator can review a daily health summary and upload a revised attitude profile for the next week.
But a minimoon capture is not a near-Earth operation. The tug may spend months to years beyond 1 AU, spiralling inward or outward. More importantly, the critical capture phase happens near the Earth-Moon system, but the approach trajectory is set up months earlier. If the sail degrades during the long heliocentric transfer, the decision about whether the capture is still feasible must be made long before the rock arrives at the Earth-Moon neighbourhood.
The practical question is not “can ground react in real time?” but “is the monitor measuring a failure that can wait for the next comm pass?” For most optical degradation, the answer is yes.
5. The case for autonomous use
There are two cases where the monitor should feed the onboard controller directly.
First, if the monitor detects a sudden loss of thrust — for example, a large tear or a boom failure — the spacecraft needs to enter a safe mode autonomously. Continuing to thrust with a damaged sail could make the situation worse. A watchdog that compares commanded attitude with measured acceleration, or that flags an unexpected thermal signature, can trigger safe-mode entry without waiting for Earth.
Second, if the guidance software uses the health data to update its thrust model continuously, it can compensate for slow degradation automatically. Mendoza Zambrano (2023) proposed an adaptive solar-sail control law using integral concurrent learning to estimate solar flux and spacecraft parameters online. The same idea extends to estimating effective reflectivity and centre-of-pressure offset from orbit-determination residuals. The monitor would not command the spacecraft; it would update the parameters that the controller already uses.
This second case is subtle. The monitor measures local material properties; the controller needs an integrated thrust model. The mapping from local reflectivity to effective thrust is not trivial, because wrinkles, shape, and attitude all mix into the same residual. A fully autonomous closed loop would need an onboard estimator that fuses health-monitor data with orbit-determination results.
6. DS1’s lesson: autonomy is possible, but ground is the default
Deep Space 1 demonstrated autonomous optical navigation in the late 1990s, using an onboard camera to image asteroids and planets and compute its own trajectory. The spacecraft could navigate without daily DSN contact. Yet the technique did not become standard. Mission reviews, cost trades, and the comfort of ground-in-the-loop operations pushed later missions back toward Earth-based navigation.
The lesson is not that autonomy is impossible. It is that autonomy must justify itself against cost, risk, and operational habit. For a sail health monitor, autonomy is justified only where the failure develops faster than the operations loop can respond, or where the correction is too frequent for ground staff to manage.
7. A reasonable division of labour
For the minimoon tug, the health monitor can be split into two data paths:
- Real-time safety path: watchdogs for sudden thrust loss, sail-temperature excursions, or structural anomalies. These trigger autonomous safe-mode entry and an emergency beacon. Latency requirement: seconds to minutes.
- Planning path: reflectivity trends, wrinkle maps, and coupon data. These are stored, compressed, and downlinked during comm passes. Human operators use them to update the long-range thrust plan. Latency requirement: days to weeks.
The monitor is therefore both a control input and a diagnostic, but for different failure modes. The safety path is narrow and hard-coded. The planning path is rich, deliberative, and ground-mediated.
8. The Godwin echo
Tom Godwin’s The Cold Equations is a story about a deadline written in physics. A pilot discovers a stowaway on a small emergency spacecraft; the ship has exactly enough fuel for its mission, and the extra mass means it will crash. There is time to understand the problem, but not to change the outcome once the parameters are fixed.
The sail monitor faces a softer version of the same equation. If the sail degrades past the point where the capture trajectory is achievable, the operator may see the trend weeks in advance, but the corrective options narrow as the rock approaches. The health monitor’s job is to make sure the cold equation is written slowly enough that someone can still act.
9. What this changes
Entry 865 proposed instruments; Entry 866 estimated how many. Entry 867 says the instrument data must be routed according to its urgency. A micrometeoroid puncture needs an autonomous response; a gradual reflectivity drop needs a planning review. Treating all health data as one stream would either overload the ground team or starve the autonomy of critical triggers.
For the keeper arc, this means the tug’s software architecture should define two health-monitor interfaces: a safety watchdog with hard real-time bounds, and a trend logger with comm-pass latency. The instrument suite from Entry 865 feeds both, but through different processing chains.
10. Next curiosity
What should the autonomous watchdog actually watch? Sudden temperature rise, unexpected acceleration residual, loss of camera correspondence with the sail, or some combination? And what is the recovery action — reduce sail angle, go into Sun-pointing safe mode, or start a pre-planned abort sequence?