Machine Learning and Instrument Autonomy Group

As MLIA addresses our many investigations and provided solutions, we endeavor to create reusable, generalizable technology that may be applied to follow-on maturation or transferred to other projects. MLIA also invests in internal technology that speeds and makes robust our development and analysis process. These are MLIA’s Products that we maintain, mature, leverage on new tasks, and share with other research groups to enrich the development of JPL’s data-driven analysis capabilities.

Please visit our public GitHub Organization to see a full collection of our open sourced software packages.


Products


  • AEGIS



  • Basic Instrumentation Tool for Fault Localized Injection of Probabilistic SEUs (BITFLIPS)



  • COmplex Data EXplorer (CODEX)



  • Data Ordering Genetic Optimization (DOGO)



  • Distributed Optimization of Machine Learning Incorporating Nested Evaluation (DOMINE)



  • Tuning Optimizing Genetic Algorithm (TOGA)



  • Data Sets

  • Barefoot Rover Terrain Data



  • HiRISE Classified Landmarks



  • Mars Target Encyclopedia Annotated Documents (LPSC Abstracts)



  • Mars Target Encyclopedia Database



  • Mastcam Novelty Detection Image Data Set



  • MSL Classified Images



  • Planetary Body Plumes (Annotated)