NASA is flying Roman's coronagraph to learn what still breaks in orbit

Margaret Turnbull's team has already decided what a broken instrument will look like. The plan is to aim Roman's coronagraph at a handful of exoplanets already known to be in place, and if nothing comes back, the fault sits in the hardware rather than in the sky ^[18]. Turnbull, an exoplanet scientist at the SETI Institute who leads one of the coronagraph science teams, says the planets are there, so an empty image is a diagnosis ^[19]. The launch window is about to open ^[1]. Roman's main instrument is the one expected to feed practically every area of astrophysics ^[2]; the coronagraph is the experiment riding along, attempting to capture starlight reflected off a planet's surface directly for the first time ^[3]. That known-answer test is the transferable part of this mission for anyone shipping a capability they know is unfinished: without a target whose answer is already settled, a weak result could be blamed on the instrument or on the sky, and nothing would settle which. The hardware being characterised is small and specific. Two mirrors that fit in a palm each carry roughly 2,300 actuators that expand under a jolt of electricity, reshaping the surface to reverse interference ^[12], which works out to about 4,600 individually driven points of adjustment across the pair ^[22]. NASA has never flown active deformable mirrors in space before ^[13]. They need detectors sensitive enough to amplify individual photons, because the instrument will catch very few from any planet ^[14], plus a set of masks that Bruce Macintosh of the University of California Observatories calls beautiful, complicated shapes, against Hubble's, which he describes as a little piece of metal that gets in the way of the star ^[15][16]. Coronagraphs themselves are not new in orbit; Hubble and JWST both carry one ^[6]. The difference is the adaptive optics that deform the mirror to cancel distortions ^[7], standard kit at the Very Large Telescope in Chile and at Keck in Hawaii, where it corrects for the atmosphere ^[10]. Above the atmosphere there was never a reason to carry it, and no space telescope has tried to see old, cool planets lit only by reflected starlight ^[11]. The ground record, then, is background rather than validation: it is a different job done with borrowed parts, at a tolerance Turnbull describes as any little bit of starlight in the wrong place destroying a whole portion of the image ^[9]. Wired's headline calls the instrument a hidden technological leap ^[23]. Vanessa Bailey, the JPL astrophysicist serving as instrument scientist for the coronagraph, describes the same hardware as something they will test in space for the first time in order to understand what work still is left to go ^[5]. Both readings hold. The eventual goal, a glimpse of an Earthlike planet around a sunlike star, is beyond current engineering ^[4], NASA's own analogy for the task is photographing a firefly next to a floodlight from across the country ^[8], and the Habitable Worlds Observatory the agency hopes to launch perhaps in the 2040s would need a coronagraph up to 100 times more effective ^[21]. Roman's job here is narrower and more specific than closing that gap: it measures it. The mission also protects itself against the hardest case failing. The same instrument will observe stars ringed by dust and debris, where gaps can betray planets nobody has detected and the amount of clutter tells astronomers how ordinary our own solar system is ^[20]. What makes this particular test worth running is what Turnbull and Bailey have already built into it: a target whose answer is known, so a null result identifies the coronagraph's own defect rather than opening a debate, and a named recipient for that defect list, the Habitable Worlds Observatory team, whose successor design cannot be finalized until Roman reports back.