Sreeja Nag

SPHERES Zero Robotics software development: Lessons on crowdsourcing and collaborative competition

Crowdsourcing is the art of constructively organizing crowds of people to work toward a common objective. Collaborative competition is a specific kind of crowdsourcing that can be used for problems that require a collaborative or cooperative effort to be successful, but also use competition as a motivator for participation or performance. The DARPA InSPIRE program is using crowdsourcing to develop spaceflight software for small satellites under a sub-program called SPHERES Zero Robotics - a space robotics programming competition. The robots are miniature satellites, called SPHERES, that operate inside the International Space Station (ISS). The idea is to allow thousands of amateur participants to program using the SPHERES simulator and eventually test their algorithms in microgravity. The entire software framework for the program, to provide the ability for thousands to collaboratively use the SPHERES simulator and create algorithms, is also built by crowdsourcing. This paper describes the process of building the software framework for crowdsourcing SPHERES development in collaboration with a commercial crowdsourcing company called TopCoder. It discusses the applicability of crowdsourcing and collaborative competition in the design of the Zero Robotics software infrastructure, metrics of success and achievement of objectives.

Observing System Simulations for Small Satellite Formations Estimating Bidirectional Reflectance

Abstract The bidirectional reflectance distribution function (BRDF) gives the reflectance of a target as a function of illumination geometry and viewing geometry, hence carries information about the anisotropy of the surface. BRDF is needed in remote sensing for the correction of view and illumination angle effects (for example in image standardization and mosaicing), for deriving albedo, for land cover classification, for cloud detection, for atmospheric correction, and other applications. However, current spaceborne instruments provide sparse angular sampling of BRDF and airborne instruments are limited in the spatial and temporal coverage. To fill the gaps in angular coverage within spatial, spectral and temporal requirements, we propose a new measurement technique: use of small satellites in formation flight, each satellite with a VNIR (visible and near infrared) imaging spectrometer, to make multi-spectral, near-simultaneous measurements of every ground spot in the swath at multiple angles. This paper describes an observing system simulation experiment (OSSE) to evaluate the proposed concept and select the optimal formation architecture that minimizes BRDF uncertainties. The variables of the OSSE are identified; number of satellites, measurement spread in the view zenith and relative azimuth with respect to solar plane, solar zenith angle, BRDF models and wavelength of reflection. Analyzing the sensitivity of BRDF estimation errors to the variables allow simplification of the OSSE, to enable its use to rapidly evaluate formation architectures. A 6-satellite formation is shown to produce lower BRDF estimation errors, purely in terms of angular sampling as evaluated by the OSSE, than a single spacecraft with 9 forward-aft sensors. We demonstrate the ability to use OSSEs to design small satellite formations as complements to flagship mission data. The formations can fill angular sampling gaps and enable better BRDF products than currently possible.

Cost and risk analysis of small satellite constellations for earth observation

Distributed Space Missions (DSMs) are gaining momentum in their application to Earth science missions owing to their ability to increase observation sampling in spatial, spectral, temporal and angular dimensions. Past literature from academia and industry have proposed and evaluated many cost models for spacecraft as well as methods for quantifying risk. However, there have been few comprehensive studies quantifying the cost for multiple spacecraft, for small satellites and the cost risk for the operations phase of the project which needs to be budgeted for when designing and building efficient architectures. This paper identifies the three critical problems with the applicability of current cost and risk models to distributed small satellite missions and uses data-based modeling to suggest changes that can be made in some of them to improve applicability. Learning curve parameters to make multiple copies of the same unit, technological complexity based costing and COTS enabled small satellite costing have been studied and insights provided.

Relative trajectories for multi-angular earth observation using science performance optimization

Distributed Space Missions (DSMs) are gaining momentum in their application to earth science missions owing to their unique ability to increase observation sampling in spatial, spectral and temporal dimensions simultaneously. This paper identifies a gap in the angular sampling abilities of traditional monolithic spacecraft and proposes to address it using small satellite clusters in formation flight. The science performance metric for the angular dimension is explored using the Bidirectional Reflectance-distribution Function (BRDF), which describes the directional variation of reflectance of a surface element. Previous studies have proposed the use of clusters of nanosatellites in formation flight, each with a VNIR imaging spectrometer, to make multi-spectral reflectance measurements of a ground target, at different zenith and azimuthal angles simultaneously. In this paper, a tradespace of formation flight geometries will be explored in order to optimize or maximize angular spread and minimize BRDF estimation errors. The simulated formation flight solutions are applied to the following case studies: Snow albedo estimation in the Arctic and vegetation in the African savannas. Results will be compared to real data from previous airborne missions (NASAs ARCTAS Campaign in 2008 and SAFARI Campaign in 2000).

Evaluation of Hyperspectral Snapshot Imagers onboard Nanosatellite Clusters for Multi-Angular Remote Sensing

Hyperspectral snapshot imagers are capable of producing 2D spatial images with a single exposure at selected and numerous wavelength bands instead of 1D spatial at all spectral band images like in push-broom instruments. Snapshot imagers are critical technologies for multi-angle remote sensing using distributed space missions. They help to relax the attitude control requirements of clusters of small satellites whose narrow field-of-view payloads point at the same ground spot or to increase the footprint area of small satellite constellations with wide field-of-view payloads. This paper reviews the existing spectral imagers for multi-angle remote sensing, performs a feasibility study to incorporate existing state-of-the-art snapshot imagers and proposes baseline imagers to serve as payload for the distributed nanosatellites. The overall approach includes an extensive trade study to identify the optics, spectral elements, their parameters and compare the identified choices both qualitatively and quantitatively. The proposed baseline design has an telescope aperture diameter of 7 cm, focal plane pixel size of 20 μm, 1000 pixels per side of the focal plane array sampling the scene and acousto-optic tunable filters or waveguide spatial heterodyne imagers that simulate a swath up to 90 km, image up to 86 wavebands with an SNR above 100. The tradeoff between spectral and spatial ranges sampled by the two baseline imager options has been highlighted.

Laser communication system design for the Google Lunar X-Prize

Laser communications systems offer a significant advantage over traditional radio frequency systems due to the shorter wavelength of laser light. Data can be sent at higher rates for less power with proportionally smaller transmitters and receivers. A laser communications system has never been demonstrated in a moon to Earth link at MBps data rates This paper presents a model of a laser downlink from the moon, developed to rapidly explore various system architectures. Modeling and analysis shows that the target data rates of above 2 Mb/s are possible with 300mW of transmitter power using a 3.5 mm aperture, a 1.5 m receiver diameter and a minimum gimbal resolution (maximum step size) of 78 µrad. 19.89 Mb/s data rates are possible using a 1 cm transmit aperture but with a much stricter minimum gimbal resolution (maximum step size) of 27 µrad.

Multispectral Snapshot Imagers Onboard Small Satellite Formations for Multi-Angular Remote Sensing

Multispectral snapshot imagers are capable of producing 2-D spatial images with a single exposure at selected, numerous wavelengths using the same camera, therefore, operate differently from push broom or whiskbroom imagers. They are payloads of choice in multi-angular, multi-spectral imaging missions that use small satellites flying in controlled formation, to retrieve Earth science measurements dependent on the target’s bidirectional reflectance-distribution function. Narrow fields of view are needed to capture images with moderate spatial resolution. This paper quantifies the dependencies of the imager’s optical system, spectral elements, and camera on the requirements of the formation mission and their impact on performance metrics, such as spectral range, swath, and signal-to-noise ratio (SNR). All variables and metrics have been generated from a comprehensive, payload design tool. The baseline optical parameters selected (a diameter of 7 cm, a focal length of 10.5 cm, a pixel size of

Subsystem support feasibility for formation flight measuring Bi-directional Reflectance

20~\mu \text{m}

Gross primary productivity estimation using multi-angular measurements from small satellite clusters

, and a field of view of 1.15°) and snapshot imaging technologies are available. The spectral components shortlisted were waveguide spectrometers, acousto-optic tunable filters (AOTF), electronically actuated Fabry–Perot interferometers, and integral field spectrographs. Qualitative evaluation favored AOTFs, because of their low weight, small size, and flight heritage. Quantitative analysis showed that the waveguide spectrometers perform better in terms of achievable swath (10–90 km) and SNR (>20) for 86 wavebands, but the data volume generated will need very high bandwidth communication to downlink. AOTFs meet the external data volume caps well as the minimum spectral (wavebands) and radiometric (SNR) requirements, therefore, are found to be currently feasible and design changes to improve swath suggested.

The power of inexpensive satellite constellations

Distributed Spacecraft Missions can be used to improve science performance in earth remote sensing by increasing the sampling in one or more of five dimensions: spatial, temporal, angular, spectral and radiometric. This paper identifies a gap in the angular sampling abilities of traditional monolithic spacecraft and proposes to address it using small satellite clusters in formation flight. The angular performance metric chosen to be Bi-directional Reflectance Distribution Function (BRDF), which describes the directional and spectral variation of reflectance of a surface element at any time instant. Current monolithic spacecraft sensors estimate it by virtue of their large swath (e.g. MODIS, POLDER), multiple forward and aft sensors (e.g. MISR, ATSR) and autonomous maneuverability (e.g. CHRIS, SPECTRA). However, their planes of measurement and angular coverage are limited. This study evaluates the technical feasibility of using clusters of nanosatellites in formation flight, each with a VNIR (visible and near infra-red) imaging spectrometer, to make multi-spectral reflectance measurements of a ground target, at different zenith and azimuthal angles simultaneously. Feasibility is verified for the following mission critical, inter-dependent modules that need to be customized to fit specific angular and spectral requirements: cluster geometry (and global orbits), guidance, navigation and control systems (GNC), payload, onboard processing and communication. Simulations using an integrated systems engineering and science evaluation tool indicate initial feasibility of all listed subsystems.