Join AeroVironment to build machine learning and computer vision capabilities that support real-time automated target recognition for both defense and commercial use cases. This on-site role in Centreville, VA includes hands-on experimentation, validation/verification testing, and opportunities to support fielded systems with extended travel. You’ll work within the Learning and Active Perception (LEAP) group, from requirements gathering through deployment-focused support.
Compensation: USD 102,000 - 155,500 per year. Work arrangement: 9/80 schedule.
Responsibilities
- Develop computer vision and ML algorithms for object detection, classification, localization, and tracking across stationary and mobile sensor platforms, focused on real-time automated target recognition (ATR)
- Conduct visual imagery data science to guide data collection, data labeling, and data selection for training deep computer vision models
- Train algorithms and validate both data selection and algorithm design through purpose-designed experiments
- Analyze ML approaches for a given problem and rank models by success probability on new data
- Set objectives, develop models, and define metrics to track progress
- Select appropriate datasets and data representations
- Perform error analysis across data, model, and design strategies, then apply corrective approaches
- Write and test software to integrate ML algorithms into aircraft components (such as autopilots, payloads, or other functional systems) or other systems
- Other duties as assigned
Requirements
- BS in Computer Science/Engineering, Mathematics, Physics, or a related Science or Engineering discipline, with 2+ years of relevant machine learning and/or computer vision experience
- 2 - 5 years of relevant computer vision experience
- Familiarity with Python and C/C++
- Proficiency with a deep learning framework, preferably PyTorch
- Proficiency with basic ML libraries such as Open-CV, scikit-learn, and pandas
- Familiarity with Linux
- Familiarity with Nvidia Tools (CUDA, JetPack, TensorRT) and the deployment process to Nvidia GPUs
- Demonstrated ability to troubleshoot complex systems and perform algorithmic optimization
- Ability to perform exploratory data analysis, including visualizing and manipulating large datasets
- U.S. Citizenship required
- Active Secret level (or higher) DoD security clearance
- Willingness and ability to travel CONUS and OCONUS (Occasionally, up to 25%)
- Location: Centreville, VA or Minneapolis, MN
Technologies
Python, C/C++, PyTorch, Open-CV, scikit-learn, pandas, Linux, Nvidia Tools (CUDA, JetPack, TensorRT), Nvidia GPUs
Benefits
- Medical, dental, and vision coverage
- 401K with company matching
- 9/80 work schedule
- Paid holiday shutdown
Clearance, ITAR, and Eligibility
This role requires access to information subject to ITAR and/or EAR. Applicants must qualify as a U.S. person under ITAR/EAR, or be approved for an export license by the governing agency. An offer that requires export license approval is conditional on AeroVironment’s ability to obtain the license in a timeframe aligned with business needs.
U.S. Citizenship is required. Secret or Top Secret clearance, or the ability to obtain a clearance, is desired.
Physical demands: Ability to work in an office environment (constant). Required to sit and stand for long periods; talk, hear, and use hands and fingers to operate a computer keyboard (frequent).
