Introduction
‘Oumuamua (1I/2017 U1) was the first unambiguously interstellar object to be discovered in the solar system (Meech et al., Reference Meech2017). Controversially, it has been suggested to be an alien spacecraft (Bialy and Loeb, Reference Bialy and Loeb2018; Loeb, Reference Loeb2021). Alien activity as an explanation for astronomical phenomena is often hard to follow up as these suggestions tend not to produce testable predictions. If researchers cannot see what research programme an idea points to, then they have no incentive to change from their present research to this new field (Lakatos, Reference Lakatos1980). This paper points out that a well-defined research programme does arise for an alien craft interpretation of ‘Oumuamua, leading to tests of this hypothesis that have various degrees of difficulty. This approach in no way rejects a natural explanation for ‘Oumuamua, but simply aims at an impartial examination of the alien craft hypothesis by following the questions arising a few more steps. Though some estimates are made here, properly quantifying all the tests will require engaging a wide range of expertise.
The implications for the limited number of possible scenarios that flow from the assumption that ‘Oumuamua is an alien craft are explored below. Katz (Reference Katz2021) has made some similar points, albeit in a less open-ended fashion. The same general considerations will apply to any future objects for which this hypothesis is raised.
The argument in favour of the alien craft hypothesis can be readily summarized. ‘Oumuamua clearly accelerated on its way out of the solar system, a measurement made at high confidence (Micheli et al., Reference Micheli2018, but see Rafikov, Reference Rafikov2018; Katz, Reference Katz2019). The acceleration was small, ~5 × 10−6 m s−2, but highly statistically significant (30σ, Micheli et al., Reference Micheli2018). The acceleration was not constant but dropped off with a heliocentric distance close to a r−2 law. No cometary dust (Meech et al., Reference Meech2017) or gas (for CO and CO2, Trilling et al., Reference Trilling2018) ejection was observed that could cause the observed non-gravitational radial acceleration. If no mass was lost, then the remaining possibility is that ‘Oumuamua was accelerated by solar radiation pressure and the solar wind. This is prima facie plausible as the large (1.5 mag, a factor ~4) amplitude of the ‘Oumuamua light curve implies a highly elongated or highly flattened geometry, a cigar or a pancake with a large, but not unprecedented, 6:1 axial ratio (Meech et al., Reference Meech2017; Trilling et al., Reference Trilling2018; Mashchenko, Reference Mashchenko2019). A large area for intercepting solar radiation may then be available. This alone does not imply a non-natural explanation. However, to achieve the observed acceleration using these weak forces ‘Oumuamua would have to have a huge area/mass ratio ~0.1 g cm−2, due to either low density, ~10−5 g cm−3, or extreme thinness, ~0.3–0.9 mm, giving a resultant mass of only ~750 kg for this ~100 m scale object (Bialy and Loeb, Reference Bialy and Loeb2018). An extremely thin shape is suggestive of an artefact to some researchers.
Natural explanations are not at all ruled out. Several astrophysical suggestions have been made to explain the unusual shape, high area/mass ratio, and acceleration of ‘Oumuamua. e.g. a fractal icy aggregate with a density of ~10−5 g cm−3 (Moro-Martin, Reference Moro-Martin2019; Luu et al., Reference Luu, Flekkøy and Toussaint2020); a comet fragment (Sekanina, Reference Sekanina2019); molecular hydrogen ice (Seligman and Laughlin, Reference Seligman and Laughlin2020); molecular nitrogen ice (Jackson and Desch, Reference Jackson and Desch2021); a fragment from a planetoid destroyed in a tidal disruption event around a white dwarf (Rafikov, Reference Rafikov2018); and some have been critiqued (Levine et al., Reference Levine, Cabot, Seligman and Laughlin2021; Hoang and Loeb, Reference Hoang and Loeb2020; Phan et al., Reference Phan, Hoang and Loeb2021; Loeb, Reference Loeb2021; Siraj and Loeb, Reference Siraj and Loeb2021).
While most scholars find the astrophysical options plausible, for some researchers the evidence against the proposed astrophysical explanations is compelling. For these researchers, the only remaining possibility is that ‘Oumuamua is dominated by a thin, large area, ‘solar sail’ constructed by an alien civilization (Loeb, Reference Loeb2021). This proposal has itself been highly criticized (e.g., Katz, Reference Katz2021).
The approach taken here is to investigate the consequences of an alien spacecraft explanation so that these consequences can be studied further to either reinforce or reduce the likelihood of this option. The number of possibilities for this option appears to be quite limited, allowing the analytic approach used here.
Unknown technologies or physics for the alien craft are not invoked here, as that approach is unconstrained. Instead, the assumption is that the putative alien designers worked with the same physics we have, the same fundamental constraints on materials that we have, and design constraints that are familiar to us. The result is that, for any of the cases explored, their craft must perform to some demanding and sometimes well-constrained capabilities.
In the alien spacecraft hypothesis either ‘Oumuamua is piloted (not necessarily by alien life directly but by their computers) and so is under control, or it is derelict, i.e. not under control. The long operational lifetime of ‘Oumuamua in this hypothesis raises further issues. All three areas have consequences that are explored here.
Un-Guided or derelict
The simpler of the two possibilities is that ‘Oumuamua is un-guided. ‘Oumuamua was found to have a rotation period of ~7 h, though the light curve did not repeat precisely indicating a tumbling motion (i.e. non-principal axis rotation, Meech et al., Reference Meech2017; Drahus et al., Reference Drahus2018; Fraser et al., Reference Fraser, Pravec, Fitzsimmons, Lacerda, Bannister, Snodgrass and Smolić2018). This rotation state suggests that it is not under control, or not wholly under control. The object was then behaving like a derelict vessel. ‘Oumuamua entered the solar system with a velocity very close to that of the Local Standard of Rest (LSR, Mamajek, Reference Mamajek2017). As ‘Oumuamua is in the LSR it is more correct to say that the Sun travels at this speed in the LSR. Any option requiring ‘Oumuamua to be under control is weakened, given this drifting trajectory.
Anonymous METI
‘Oumuamua may not be truly derelict but may instead be a passive ‘message in a bottle’ (Loeb, Reference Loeb2018). In this hypothesis ‘Oumuamua was designed to let technological life (because telescopes were needed to detect ‘Oumuamua) around another star realize the existence (at least at one time) of another civilization. The solar sail, in this case, may have been intended not for realistic propulsion but to give a small but tell-tale acceleration. That would require that the sail orient itself at times perpendicular to the radial direction from the Sun. A tumbling motion will reduce the amount of time the sail is oriented correctly to achieve a significant acceleration. If the sail were always oriented radially then ‘Oumuamua would have decelerated on approaching the Sun. That possibility would affect the apparent incoming direction which could modestly affect the arguments in Sec.3.1 below.
An additional possible motivation to situate a spacecraft in the LSR would be to deliberately hide its origin. The message in the bottle would then be ‘You are not alone, but we are not telling you where we are.’ Such a design may have been motivated as being a safe version of messaging extraterrestrial intelligence (‘Anonymous METI’Footnote 1) In this case we would expect that a mission to ‘Oumuamua, or another suspiciously behaving interstellar object (e.g. Hibberd and Hein, Reference Hibberd and Hein2021), would find clear evidence of these bodies being a deliberate construct, but would have no identifying marks equivalent to the plaques mapping out Earth's location with respect to nearby pulsars on Pioneers 10 and 11 and Voyagers 1 and 2Footnote 2.
Inadvertent METI
The upper stages of the launchers that put our space probes onto their final trajectories also reach a similar trajectory. There are numerous upper stages in heliocentric orbits and a small number on solar system escape trajectories, e.g. the New Horizons upper stage. These are inadvertent interstellar spacecraft, and so a form of unintended METI from the sender's perspective. From our perspective, they are another example of a technosignature.
Similarly, the putative builders of ‘Oumuamua may have manufactured these craft as upper stages, presumably accelerated by directed energy (e.g. lasers), given their light-sail like design. This form of acceleration is a good way to overcome the limits imposed by the rocket equation. In this case, there must have been a highly active stellar system economy around the origin star ~105 year ago. Possibly this activity would leave other tracers, for example from the mining of asteroids for the raw material from which to build the craft (Forgan and Elvis, Reference Forgan and Elvis2011). These ‘craft’ may even be just tailings from mining in the stellar system's Oort cloud. These tailings would have been given a small velocity to expel them into interstellar space, perhaps in order to reduce threats to their infrastructure in their Oort cloud. Very approximately, comets at 104 AU will be moving at ~30 m s−1, sufficient to damage loosely bound objects (including spacecraft). The space density of Oort cloud comets from 104–105 AU is only ~1 per 4000 AU3 making collisions rare, but if say 10% of the comets were cut into ~1 t slices there would be ~1021 of them (Weissman, Reference Weissman1983) for a space density of ~105 AU−3 a separation of ~5 × 106 km, that they would traverse in ~5 years. Though their cross-section is ~0.01 km2, and their probability of collision ~2 × 10−15 there would still be ~107 collisions/year. For a civilization with large amounts of infrastructure in their Oort cloud, this could be sufficient motivation to eject the comet slices. A true runaway ‘Kessler Syndrome’ (Kessler and Cour-Palais, Reference Kessler and Cour-Palais1978) may not develop.
The ‘message’ from inadvertent METI would be simply ‘we exist’ with any other information gleaned from a close inspection being also inadvertent.
Population implications
As ‘Oumuamua has a ~105–106 year travel time (see Sec.3.1), it was constructed before any detectable human technology was created, possibly even before homo sapiens evolved. This implies that the ‘message in a bottle’ strategy must have been implemented without knowing which stars would have technological societies capable of detecting ‘Oumuamua. There must then be at least a moderately large number of copies in order to swing through the nearby solar systems at a rate that makes their detection likely given distribution of onset times and lifetimes for this capability. The LSR velocity argues against efficient targeting, keeping the total number large.
If ‘Oumuamua came into the solar system entirely by chance, without even the ‘message in a bottle’ plan behind it, then the population density of such objects, at least locally, would be the same as if it were a naturally occurring object, and interstellar asteroid. On this basis Do et al. (Reference Do, Tucker and Tonry2018) estimate the space density of ‘Oumuamua-like objects to be 0.2 AU−3 or 2 × 1015 pc−3. Loeb (Reference Loeb2018) says that ‘This would require the unreasonable rate of a launch every five minutes from a planetary system even if all civilizations live as long as the full lifetime of the Milky Way galaxy.’ This high rate has been taken as arguing for a natural origin.
The maximum total mass of all copies of ‘Oumuamua in this scenario would be ~1014 t (metric tonnes, for a density ~10−4 g cm−3). While far larger than the Earth's proven iron reservesFootnote 3 (2.3 × 1011 t), this mass is small compared with, e.g. the ~1018 t of iron in the Main Belt asteroids (Elvis and Milligan, Reference Elvis and Milligan2019), and deployment could well have taken place over 1% of the planned 105 year lifetime, or ~103 year, as this implies a launch rate of ~3000 s−1. For an Earth-scale economy, this is an unattainable number, but for a solar system scale economy, where the Main Belt asteroids are fully exploitable, many thousands of ‘Oumuamua factories and launchers are within the range of plausibility.
The putative designers may well have known which planets harboured, or could harbour, complex life. As roughly a quarter of all M-dwarfs have rocky planets in their habitable zones (Dressing and Charbonneau, Reference Dressing and Charbonneau2015), this knowledge may not reduce the number of alien craft required by a large factor. As we learn more about the habitability of exoplanets the fraction may go down. For example, stellar activity may remove the atmospheres of all Earth-sized planets in the habitable zones of M stars (Van Eylen et al., Reference Van Eylen, Agentoft, Lundkvist, Kjeldsen, Owen, Fulton, Petigura and Snellen2018) The total numbers and the required launch rate may then be reduced by some orders of magnitude if the putative builders had this knowledge.
Ongoing and expanded surveys (e.g. VRO/LSSTFootnote 4, Moro-Martín et al., Reference Moro-Martín, Turner and Loeb2009) for more interstellar asteroids will determine the population density of interstellar objects of whatever kind and so put constraints on this case.
Under control
If ‘Oumuamua was under control during the solar system transit, then all the properties of ‘Oumuamua are deliberate and require investigation. The first question is why is it in the LSR if the Sun is its destination? Additional questions immediately arise about (1) the origin and (2) the destination of ‘Oumuamua, and (3) the achievable accuracy of the inbound trajectory. As the design is dominated, in this hypothesis, by a solar sail, the craft was surely using that solar sail to navigate when near a radiation source.
Where did it come from?
‘Oumuamua must be coming from some other star if it uses the light pressure from that star to accelerate. As the light sail dominates the system mass then to be justify that mass, the origin star ought to be nearby, so that ‘Oumuamua does not spend most of its time in the isotropic illumination of interstellar space, where the solar sail is of no use. Active reflectivity control of the sail could continue to generate a modest acceleration. Otherwise, the use of a light sail would be a poor engineering choice. In addition, the challenge of lifetimes >106 years are significant, especially for large thin structures (see section 4) arguing for a local origin. There are several possibilities for the origin star.
‘Oumuamua had a hyperbolic orbitFootnote 5 (e = 1.20), arriving from (RA, dec) = (18 h 42 m, +34.3o, ± 5′), with an asymptotic velocity of 26 km s−1 (Meech et al., Reference Meech2017; Bailer-Jones et al., Reference Bailer-Jones2018). At this speed ‘Oumuamua travels 2.5 pc in 105 years.
Main sequence stars
No catalogued stars lying within 60 pc of the Sun come within one parsec of the trajectory, and only four stars come within 4 pc in SIMBAD (Dybczyński and Królikowska, Reference Dybczyński and Królikowska2018), and only 6 stars have come within 2 pc in the deeper Gaia DR2 catalogue over the past 106 year (Bayler-Jones et al., Reference Bailer-Jones2018). These papers assume no manoeuvres before the detection of ‘Oumuamua. The minimum course corrections needed to create a rendezvous with another star in the past would be useful to calculate.
Only stars with main-sequence lifetimes of ~1 Gyr or more are likely to have developed complex life. By terrestrial standards, 1 Gyr is barely enough time to harbour multi-cellular life. The oldest certain fossils found so far on Earth come from a time when the Earth was ~1 Gyr old (Schopf et al., Reference Schopf2017). There are suggestions of single-cell life as early as 0.3 Gyr (Dodd et al., Reference Dodd2017). Complex life, however, seems to have taken much longer, ~4 Gyr, although the Cambrian explosion that happened then was rapid (Parry et al., Reference Parry, Boggiani, Condon, Garwood, JdeM Leme, McIlroy, Brasier, Trindade, Campanha, Pacheco, Diniz and Liu2017). Hence, stars with masses >2.5 M sol that have Main Sequence lifetimes <1 Gyr, are unlikely origin stars (spectral type A2 or earlier). This concern, though, assumes that Earth is a typical example, which it may not be as we do not know what triggered the Cambrian explosion.
Brown dwarfs
Could ‘Oumuamua have come from an undetected star within ~10 light-years? The obvious, and quite numerous, population of quite hard to detect stars would be brown dwarfs, i.e. T and L stars. T and L stars are long-lived enough to potentially develop life in their systems.
These stars have been searched for in infrared surveys with UKIDSS and WISE, and in deep optical imaging for high parallax and proper motion (Best et al., Reference Best, Liu, Magnier and Dupuy2021). None presently lie within 1o of the origin point of ‘Oumuamua.
Brown dwarfs have surface temperatures, T ~ 1000 K, versus 6000 K for the Sun. As the emitted black body power goes as σ T4, a brown dwarf will produce just ~10−3 of the acceleration that the Sun provides. The launching process would then have to be by other means. An electric sail in one option (Lingam and Loeb, Reference Lingam and Loeb2019). A laser that reflects off the sail would be an efficient design choice, as for the Breakthrough StarshotFootnote 6 project. The energy required to accelerate the mass of ‘Oumuamua (~1 t) to ~20 km s−1 is 2 × 1011 J, so a 1 GW laser would have to operate for 200 s to achieve this effect. This is not far from the Breakthrough Starshot power and duration requirements.
Origin star went supernova
Could the star of origin have gone supernova since ‘Oumuamua departed ~105 years ago and so no longer be present?
Supernova remnants are detectable up to ~105 years after the event (e.g. Cygnus Loop at 2 × 104 year, and a radius of 18.5 pc, Fesen et al., Reference Fesen, Weil, Cisneros, Blair and Raymond2018) and so may still be visible, depending on how soon after the departure of ‘Oumuamua the supernova happened. If a neutron star was created in the supernova then that object should still be visible as a thermal source at ~106 K (Bignami, Reference Bignami1996), though it may have moved significantly in that time (~250 pc for a typical ejection speed of 250 km s−1, Faucher-Giguère and Kaspi, Reference Faucher-Giguère and Kaspi2006), in an unknown direction. Old supernova ejecta shells travel an order of magnitude faster than ‘Oumuamua (~200 km s−1, Fesen et al., Reference Fesen, Weil, Cisneros, Blair and Raymond2018) and so would have travelled well past the Sun and would now lie in the opposite direction to the ‘Oumuamua origin point and would have a large angular size.
Detection of large diameter supernova remnants via radio and UV emission has been improving. Fesen et al. (Reference Fesen2021) report three strong candidates at distances of ~100 pc. Finding convincing evidence of a ~10 pc distant supernova remnant will be challenging but could be attempted. Abundance measurements of the surfaces of asteroids might show recent enhancements of 26Al and other short-lived radioactive nuclei (with half-lives of 105–107 year) characteristic of Type II supernova (Adams, Reference Adams2010) The samples returned by Hayabusa 2 and OSIRIS-REx should be able to test this scenario. Massive stars tend to form in groups and there are only two plausible groups within 25 pc, B Pic and AB Dor (Mamajek, Reference Mamajek2015); neither is close to the origin direction of ‘Oumuamua.
A high mass star is unlikely to have been the location where the craft was built. The massive (8 < M < 40 M sol) stars that create core-collapse type II supernovae are short-lived (<3 × 107 year, Woosley et al., Reference Woosley, Heger and Weaver2002). The Earth was mostly molten for the first ~107 year after the Moon-creating impact, the Hadean Eon (Sleep et al., Reference Sleep, Zahnle and Lupu2014). It is hard to imagine life, let alone a technological civilization, growing under such conditions and so is highly unlikely within the lifetime of the star. In that case, the supernova star would have been only a way station for ‘Oumuamua.
This option raises other questions beyond the scope of this paper: how much would the supernova light and ejecta accelerate ‘Oumuamua given the sail area (see Sec. 4 of Do et al., Reference Do, Tucker and Tonry2018)? How far from the supernova would ‘Oumuamua have to be for the sail to survive the ejecta passage at velocities of 1000 km s−1 or more.
Compact objects as way stations
‘Oumuamua may not have come to us not from the star system where it was built, but instead after making a slingshot gravity assist manoeuvre around another object that was merely a way station for the craft. Why a way station would be used to deliver ‘Oumuamua into the LSR rather than a swifter trajectory is a puzzle.
Compact objects, either an isolated neutron star or a black hole, are particularly effective gravity assist tools in order to set course for the Sun.
A neutron star is a plausible candidate as the expected space density of neutron stars local to the Sun is predicted to be ~3.6 × 10−4 N9 pc−3. N9 is the total number of isolated neutron stars in the Galactic disc in units of 109, that is the current best estimate (Sartore et al., Reference Sartore, Ripamonti, Treves and Turolla2011). This implies that the most likely distance to the nearest neutron star is ~10 pc (Sartore et al., Reference Sartore, Ripamonti, Treves and Turolla2011). With some luck then an isolated neutron star could have been well-positioned to serve as a gravity assist body for ‘Oumuamua. Isolated neutron stars often acquire a high velocity from the supernova event in which they formed, ~250 km s−1 (Sartore et al., Reference Sartore, Ripamonti, Treves and Turolla2011) This makes them attractive as gravity assist targets (Johnson, Reference Johnson2003), although the LSR-like velocity of ‘Oumuamua suggests only a weak velocity boost.
Isolated black holes are estimated to be about an order of magnitude rarer than isolated neutron stars with ~108 in the Galactic disc and bulge (Fender et al., Reference Fender, Maccarone and Heywood2013). For a uniform disc distribution (a more naïve calculation than that of Sartore et al., Reference Sartore, Ripamonti, Treves and Turolla2011), Fender et al. (Reference Fender, Maccarone and Heywood2013) find a density of 10−4 pc−3, and a mean separation of just over 10 pc, so again it is plausible that an isolated black hole was conveniently located for ‘Oumuamua.
For both types of a compact object, the hypothetical trajectory designers need to have mapped out the local population of both types of object in position and velocity to high enough precision to ensure that the encounter was at the right impact parameter to produce an outbound trajectory that would give an impact parameter with the Sun of ~0.25 AU, as discussed in Sec. 3.2 below.
Both neutron stars and black holes are hard to detect when isolated. Limits can be put on both types of object in soft X-rays and in the EUV. Within ~100 light-years interstellar absorption is low, and a handful of unidentified EUV candidates exist (Maoz et al., Reference Maoz, Ofek and Shemi1997). These may be neutron stars, but none lies near the ‘Oumuamua origin point. At a typical space velocity of 250 km s−1 (Faucher-Giguère and Kaspi, Reference Faucher-Giguère and Kaspi2006), the putative neutron star of origin for ‘Oumuamua will have travelled 25 pc in 105 years, or up to 2.5° on the sky at a distance of 10 pc, depending on the direction of travel. Searches using the eROSITA all-sky survey instrument on Spektr-X-gamma (Predehl, Reference Predehl2017), and with more targeted soft X-ray and EUV instruments can be undertaken.
Bialy and Loeb (Reference Bialy and Loeb2018) estimate the tensile stress on a body due to tidal forces, P tid, will be ~10−6 r AU d42 M/M sol rAU−3 dyne cm−2, where r AU is the distance of ‘Oumuamua from the star (the Sun in the case considered by Bialy & Loeb) in AU, d 4 is the length of ‘Oumuamua in units of 104 cm, and M/M sol is the stellar mass in solar units. At an altitude of 300 km (~2 × 10−6 AU) above an M/M sol = 1 compact object, P tid ~ 1 × 1011 dyne cm−2. This is comparable to the tensile strengths of diamond or monocrystalline silicon (see Table 1 in Bialy and Loeb, Reference Bialy and Loeb2018). These theoretical limits to material strengths imply that there is a closest possible approach to both a neutron star and a black hole, which will limit the gravity assist manoeuvres achievable.
Table 1. Summary of cases, issues, and test (section numbers are in parentheses)
Course navigation
How accurately can a spacecraft be navigated using only an initial external radiation powered phase? Is that accuracy sufficient to make a precision ‘landfall’ within a fraction of an AU at a target star ~10 parsecs away? An error of 1 arcsec at 10 pc is 1.5 × 109 km (10 AU), compared with the perihelion of ‘Oumuamua of 3.7 × 107 km (0.255 AU1). To reliably put ‘Oumuamua on a solar gravity assist course the trajectory must then be accurate to the milli-arcsecond level. If instead the goal was only to get close to Earth, as a complex life-bearing planet, but avoid the Sun, then 0.01 arcsecond, or even 0.1 arcsecond, accuracy may suffice.
Solar sails employing starlight are only effective out to ~130 AU from a solar-type G star, after which the solar the radiation field becomes <1% of the diffuse Galactic optical background light (~10−9 erg s−1 cm−2 sr−1 Å−1, 0.2–1 μm, Bernstein et al., Reference Bernstein, Freedman and Madore2002) and the effective acceleration becomes negligible. Later course corrections are not possible until the radiation from the target star becomes significant, again at ~130 AU for a G star. A similar lack of course correction arises for the gravity assist case. Alternatively, the sail may be powered by a laser from farther back along the trajectory to enable a course correction. The feasibility of a laser system designed to work over parsec distances are demanding and should be investigated.
Achieving milliarcsecond accuracy in a gravity assist is particularly challenging, as the deflections gained from gravity assists are highly sensitive to the precise impact parameter of the incoming object from the massive object (e.g. Yefremov, Reference Yefremov2020). The designers would have needed to know the 3D locations and space velocities of the compact object target to high accuracy in order to achieve the correct impact parameter ~105 years later. It would be interesting to determine whether there are fundamental limits to the achievable accuracy, given even small perturbations from other objects along the path, e.g.comets, interstellar asteroids.
Course corrections could be made at a later point in the voyage, when the needed corrections could be more accurately measured. Such corrections would be correspondingly larger and would need a supplemental power source. A 1 arcsecond trajectory error can be corrected with a delta-v = 13 cm s−1 for a 26 km s−1 velocity. For ‘Oumuamua, with a mass of ~1 mt (Bialy and Loeb, Reference Bialy and Loeb2018), this would require just 16.9 J. A one degree correction of a 1 mt payload would require 6.1 × 104 J for a delta-v of 0.49 km s−1, or 100 W for 5 min. This is not a demanding requirement. For a typical chemical rocket exhaust velocity of 104 m s−1, the rocket equation gives a fuel mass of 120 kg. A rough limit to the correction would be reached when the fuel mass is equal to the payload mass, which happens when delta-v = 6.9 km s−1. Higher exhaust velocities, e.g. from ion engines, would allow larger delta-v within the mass limit. A zero propellant, albeit slow, way to change course would be for the light sail to change its albedo to be high on one side and low on the other, leading to an asymmetric force even in a uniform background (e.g. Ehresmann, Reference Ehresmann2021).
Where is it going?
‘Oumuamua is departing towards (RA, dec) = (23 h 51 m 24.3 s, +24o 42′ 59.0″) (Bayler-Jones et al., Reference Bailer-Jones2018). This exit asymptote from the solar system does not point toward any known nearby starFootnote 7 (Best et al., Reference Best2018).
‘Oumuamua will have a distant (~0.46 pc) encounter at ~100 km s−1 with the currently 3.1 pc distant M5 dwarf Ross 248 in 29 000 years (Bayler-Jones et al., Reference Bailer-Jones2018). A ~10o course correction early in the flight toward Ross 248 would bring ‘Oumuamua to a close encounter with Ross 248. The acceleration of ‘Oumuamua around the Sun (Micheli et al., Reference Micheli2018) altered the outbound direction by of order a few arcminutes compared with an extrapolation of the inbound hyperbola (Bayler-Jones et al., Reference Bailer-Jones2018). A degree sized manoeuvre was then plausibly within the capability of ‘Oumuamua, if it is an alien craft. Unfortunately, we do not now have the means to observe such a course change. Perhaps the solar encounter was a failure? This would contradict the extreme accuracies invoked earlier.
As in the case of the incoming trajectory, ‘Oumuamua could be heading towards an undetected compact object for a gravity assist. The space density of these objects is low enough that invoking one at both ends of the trajectory seems extreme. However, they can be searched for in the same way as for the incoming trajectory.
Without a destination the large deflection of ‘Oumuamua as it rounded the Sun becomes another unexplained, low probability, event, and so argues against the under-control alien craft hypothesis.
Reliability engineering considerations
Lifetime
The reliability engineering challenges (see e.g. O'Connor and Kleyner, Reference O'Connor and Kleyner2012) in designing to a 105 year lifetime are daunting. This lifetime is ~20× longer than the oldest human-built structures on Earth, most of which, though massive and built from robust materials, have suffered significant damage, albeit in a challenging environment. More relevant is that this lifetime is ~1000× longer than the lifetime of any industrial revolution machine, and ~4000× longer than the longest-lived spacecraft, Voyagers 1 and 2 launched in 1977.Footnote 8
At 10 000 years, the ‘Clock of the Long Now’ has the longest design lifetime of any current technologyFootnote 9. The lifetime of the Clock of the Long Now is a demonstration that plausible design lifetimes ~100 times greater than existing machinery can be conceived of. The lifetime needed for ‘Oumuamua if it is an alien craft is 10× longer still, which is however a relatively small factor. Studying the Clock of the Long Now principles may show ways that lifetimes comparable with the ‘Oumuamua journey time could be realized. If ‘Oumuamua came to us from a way point, rather than from the site of its construction, then all the lifetime requirements need to be at least doubled.
The challenge of reliable operation over ~105 year means that the mean time to failure (MTTF), a measure often used to analyse failures in complex systems, is ~5× longer, 500 000 years. This timescale arises from the result that for a non-repairable system to have a 99% probability of surviving to a given time (assuming Poisson failure distributions), engineers need to design for an MTTF five times longer). Hence the designers would need to have designed for an MTTF approaching 106 years for ‘Oumuamua, ~100× longer yet than for the Clock of the Long Now. Studies made for the 100 Year Starship may be helpful starting pointsFootnote 10. Failing to find solutions would argue against the alien craft hypothesis
Materials stability
The designers of ‘Oumuamua would be constrained to the same elements that we have, although they would surely have compounds novel to us. It would be interesting to see whether we could conceive of designs that could approach these lifetimes given theoretical limits on materials properties especially for large thin structures such as light sails. For example, Bialy and Loeb (Reference Bialy and Loeb2018) show that a light sail made of diamond or silicate material, with tensile strengths of ~1010 dyne cm−2 could lose about half of its mass due to interstellar dust bombardment. That is quite a narrow engineering margin, so the sail may well fail before any post-solar stellar encounter. If the craft used a gravity assist around another stellar body, then the time spent by the craft in interstellar space will have been of order a factor two longer, endangering the sail. In addition, the environment around the stellar gravity assist body may well have a higher density of dust. Similarly, encounters with zodiacal dust in the solar system would be relatively short-lived, but at higher speed and particle density. More sophisticated modelling using detailed material properties would be valuable.
Diamond sheets of this type can be made but large single-crystal sheets are well beyond our technology today (Butler and Sumant, Reference Butler and Sumant2008), though not theoretically ruled out. However, graphite is in a lower energy state than diamond and diamond will convert to graphite at raised temperatures or under ion bombardment (Setton et al., Reference Setton, Bernier and Lefrant2002). Dust impacts will raise temperatures locally (on a ~mm scale), while cosmic ray bombardment is continuous. Estimates of the lifetime of diamond sheets under the conditions of interstellar space would be instructive. In principle superconducting magnets could be used to create an artificial magnetosphereFootnote 11 to shield the light sail from cosmic rays. To keep within the mass budget these magnets must be low density and/or thin.
Silicon monocrystals are also subject to degradation at elevated temperatures and irradiation (Sperber, Reference Sperber2019), but investigations so far have been in the context of photovoltaic cells. Investigation of monocrystalline silicon under interstellar conditions would also be of interest.
Repair
For repairable systems, the appropriate metric is the less demanding mean time between failures (MTBF, O'Connor and Kleyner, Reference O'Connor and Kleyner2012). Repair raises its own issues. No outside materials would be available for ‘Oumuamua, so the craft would require recycling. Recycling has limits; material ablated off the thin sail by collisions with interstellar dust, which could be half of the initial mass (Bialy and Loeb, Reference Bialy and Loeb2018), would be lost. Repair would require a supply of raw material on board, and complex mechanisms to carry out the repairs, reducing available payload mass; the repair mechanisms themselves must have sufficiently long MTBFs. These mechanisms may have to operate after ~105 years of inaction, which presents other challenges. Repair also implies a power source during an interstellar cruise, though the amount of power may be quite small. A fault detection mechanism would seem to imply sensors and a computer that can operate for 105 years. As cosmic-ray hits accumulate with time the computer must have multiple redundancies for error avoidance and like thick connectors to survive. The necessary ‘super rad hard’ standards could be estimated.
Katz (Reference Katz2021) notes that tumbling motions will be damped out by flexing in the sail structure, and so the observed tumbling would not be expected. The sail may not flex, though, as it could have been constructed on-orbit where no deployment mechanism is required, and could then be a single (e.g. 3D printed) structure.
Keeping key elements of the craft sufficiently warm with radioisotope heating would require the use of radioactive materials with half-lives comparable to the journey time, ~105 year, greatly restricting the number of radionuclides availableFootnote 12 and requiring larger mass to obtain the same power that more rapidly decaying materials would provide. Are there alternative low-mass long-lived heat sources? Otherwise, the temperature of the craft could be allowed to be very low. A low-temperature operation may have benefits from, e.g. superconducting components, though mechanisms operating at a few K may face fundamental limits.
Discussion and conclusions
The above discussion highlights a number of challenges for the alien craft hypothesis that further analysis could examine. Table 1 summarizes all the cases, the issues they raise, and the implications and/or tests that flow from these issues. Figure 1 puts these in the context of the broader picture including astrophysical explanations.
Fig. 1. Possible families of explanations for the nature of 'Oumuamua, divided into two primary groups (astrophysical, alien craft) and sub-groups depending on properties. Possible objections and tests for each sub-group option are shown.
The ‘controlled’ case has more strikes against it than the ‘uncontrolled’ case, but neither suffers a knock-out blow, as yet. Some of the issues turn out not to be major obstacles to the alien craft hypothesis, but others cast doubt on it. All of the discussion here is highly simplified. Most of the issues suggest new, more sophisticated, studies that could be carried out. Some of these, e.g. intercept missions for newly discovered interstellar objects, are concepts already being developed and will be of value whatever these objects turn out to be.
Overall, these considerations show that a broad and well-defined research programme can be built around the hypothesis that ‘Oumuamua is an alien craft. More generally, the considerations presented here can also be applied to other interstellar visitors, as well as to general discussions of interstellar travel.
Author contributions
The author acknowledges helpful contributions from, J. Arenberg, C.J.F. Elvis, G. Fabbiano, A. Lawrence, J.C. McDowell, M. Micheli, J. Wright, and an anonymous referee.
