About the position
KBR is seeking an Intern to join its talented team in Sioux Falls, South Dakota. The Intern will be supporting the Landsat Calibration and Validation (Cal/Val) team and applying artificial intelligence and/or deep learning models to time series trended calibration data and providing automated anomaly detection. KBR plans to leverage CVAE (Convolutional Variational Auto-Encoder) and similar modelling methodologies in AWS to take an unsupervised deep learning approach to anomaly detection in current and future Landsat mission data. The overall goal for our project, 'Scaling Operations for Multi Mission Architecture', is to investigate Calibration and Validation (Cal/Val) data from the Landsat series of satellites to ensure data quality meets the gold standard in Earth observations. This includes maintaining a well-calibrated remote sensing archive of current, past, and future Landsat missions for science applications and Earth observation enhancement, as well as for Cal/Val operational trending dashboard capabilities.
Responsibilities
- Deploy baseline processing infrastructure in AWS
- Install open-source machine-learning tools: Scikit-Learn, Keras, PyTorch, etc.
- Install and explore dashboarding software (Dash, Tableau, others)
- Install and configure automated reporting software (Quarto and related software)
- Support the team with Data Preparation and Modeling
- Collect and prepare current and historical spacecraft attitude & ephemeris and other supporting data for CVAE modeling activities
- Collect and prepare a subset of the Landsat Cal/Val trending database for dashboard and automated reporting activities
- Prepare and evaluate a CVAE model
- Train and test a published CVAE model utilizing PyTorch and Scikit-Image and published data (non-Landsat) to establish a processing baseline
- Adapt the model to accept Landsat attitude & ephemeris data
- Refine and evaluate models by adjusting model parameters and augmenting the training dataset
- Investigate alternative models or modifications as needed
- Develop a test-case dashboard application in the python-based tools using a subset of the IAS trending database
- Develop a similar application in Tableau
- Evaluate functionality, compare/contrast the tools
- Participate in Internal Research and Development (IR&D) Reviews
- Perform other duties as assigned by the Landsat Cal/Val Team Lead
Requirements
- Working toward a degree in software/data science, remote sensing calibration and/or mathematics
- Experience in Machine learning and/or deep learning AI models
- Programming skills in Python, IDL, or Matlab
- Ability to communicate effectively orally and in writing
- Ability to think independently
- Ability to handle complex multitask environments
Nice-to-haves
- Experience in satellite ground systems
- Experience in remote sensing data capture, processing, archive, distribution, and scientific applications
- Knowledge of spacecraft and spacecraft instrument dynamics and characteristics
- Knowledge of image/signal processing and statistical analysis fundamentals
- Knowledge of map projections
- Ability to develop image and signal processing techniques and tools
- Knowledge of satellite systems and remote sensing
- Knowledge of cloud data processing
Benefits
- 401K plan with company match
- Medical, dental, vision, life insurance
- AD&D, flexible spending account
- Disability, paid time off
- Flexible work schedule
- Career advancement through professional training and development
