At udelv we are creating a new type of driverless vehicle. In January 2018 udelv successfully accomplished the first ever autonomous commercial “last mile” delivery made on public streets using advanced transportation and autonomous driving technologies to revolutionize this segment of transportation.
As a Perception Engineer, you will be responsible for the design, development, test and maintenance of perception software. You will have the chance to collaborate cross-functionally with other software engineers implementing autonomous driving algorithms on highly capable, next generation hardware that will allow you to explore algorithms that were traditionally too costly on edge computing systems.
- Building/Integrating software and algorithms for large scale processing of camera images, GPS/IMU, Lidar, radar and vehicle odometry
- Work with raw sensor data, low-level signal processing and analyze large datasets
- Perform multi-sensor fusion, 3D-reconstruction and multi-object tracking
- Testing algorithms in simulation, in vehicle in controlled environment, and ultimately in in vehicle in the field.
- Benchmarking and validating perception algorithms for final deployment
- Developing safety engineering and fault tolerance into perception systems.
- Collaborating with technical and non-technical team members to build full-stack autonomous solutions
- Performing other duties as necessary for completion of projects and achievement of goals.
- All other duties as assigned
- MS or PhD degree in CS/CE/EE/Robotics or equivalent with focus on Perception and Computer Vision
- 2+ years professional/ research experience in one of the following areas: multi-sensor fusion, image processing, machine learning or deep learning, and/or point cloud data processing
- Extensive experience with programming and algorithm design (C/C++, Python, Tensorflow) including ISP algorithms, DNN, CNN or point cloud processing: segmentation, classification, tracking of objects etc.
- Experience with software engineering tools and linux environment (e.g., Git, CMake, CI, gdb,etc)
- Proven ability to program real-time hardware to execute computationally intensive algorithms
- Publications in the field of Perception / Computer Vision
- Experience in Automotive Industry will be regarded. Experience in writing safety-critical code
- Familiarity with ROS and/or other message networking middleware
- SLAM / Probabilistic Filtering experience