Shane Murray Email & Phone Number

California, United States

Santa Clara, Ca, Us

Santa Clara, Ca, Us

Santa Clara, Ca, Us

Santa Clara, Ca, Us

• Tech Lead: Head of combined DNN and Algorithm teams. Company's first radar engineer. HW & SW architecture. Feature status, roadmap, and prioritization. Hiring and headcount planning. Customer & supplier management. Mentoring, reviews, & general support for team members. SW releases in NVIDIA DRIVE SDK. HW/SW releases in NVIDIA DRIVE Hyperion. Patents filed.• Algorithms: Surround object tracking and sensor fusion (IMU/odometry, cross-radar point clouds & DNN objects), adaptive statistical signal processing. Doppler ambiguity unwrapping, motion classification, and host motion compensation. Freespace for highway Lane Changes. Camera-radar fusion improvements & support.• DNN: Launch of multiple versions of DNN detectors into E2E in-car stack (objects + occupancy/freespace). Various training experiments and training/inference pipeline improvements.• Sensing: Drove HW and SW efforts from scratch for multiple generations of surround radar sensing. 77GHz & 79GHz sensor supplier & waveform selection. Sensor API & signal design. Mounting bracket and radome design guidance, RF testing, and sign off. 360 FOV architecture for low and high cost. Sensor health watchdogs. Support for HW & SW bring-up, verification, and validation.• Guidance for other radar based HW/SW such as simulation, mapping & localization, ego motion, self-calibration, and driver monitoring.• Integration with framework, sensor abstraction, rendering, perception, planning, and control. • KPI & test driven software development in C/C++, python, pytorch, linux, git, cmake, bazel, docker, and gtest. Offline bench real-world data resim, Drivesim, and many in vehicle tests and full system drives.

Stockholm, Se

(note: worked under Autoliv, which later spun off the business as Veoneer)• 24GHz Narrowband Rear Corner ADAS Radar. Product life-cycle from project sourcing to a few months from SOP. Signal processing team Scrum Master.- Embedded C programming. Offline re-sim, live sensor tests, and in-vehicle drives. - Led statistical multi-target tracking. Co-invented and patented a novel waveform based tracking technique. Validation with LIDAR, IMUs, Maps, and differential GPS. - Jointly created a complete radar signal processing chain (I/Q to Tracking) for rapid prototyping of clutter canceling, windowing, beamforming, and tracking based on real world data. - Optimized transmit waveform pulse train for blanking and range maximization. - Ran sensor calibration in RF chamber and test range. Support for online boresight auto-alignment and channel IQ calibration. - Feature support for Rear Cross Traffic, Lane Change Assist, and Blind Spot Monitoring. • 77GHz Front Radar: Ego path prediction, object lane assignment, and tracker support. Grid tracking algorithm prototype. • Completed two multi-month IEEE radar courses. • Co-invented and patented a radar camera fusion driver monitoring system.

Stockholm, Se

• Vision: Stereo and Mono Camera HW/SW setup, calibration, demos, and algorithm testing in-vehicle. Organized and drove efforts (literally) for Japan-wide vision system data collection. • Radar: 77GHz front facing radar HW/SW setup and testing. 24GHz NB radar-fascia response modelling via development of synthetic I/Q simulations based on waveform parameters, vehicle kinematics, and antenna reference frames.• Patented a cut-in detection algorithm.

Kawasaki, Kanagawa, Jp

• Designed and implemented a collision prediction algorithm using circular motion intersection via mathematical root solving. Reduced false positives from radar based Automatic Emergency Braking by 70%. Regression tested on thousands of false positive examples (Simulink/Matlab). Developed related algorithms and filed patents for host motion estimation from radar, cut-in vehicle detection, and radar ghost object suppression (JP2012-244366, JP2012-232792, JP2012-225073, JP2012-237710). • Tested a camera lane and synthetic aperture radar road fusion algorithm on bench and in-vehicle. (Simulink/MicroAutobox, C++/ADTF).• Analyzed driver attention state based on in-vehicle driving activity. Assisted testing of driver monitoring near infra red camera system. • Helped communication and cooperation between German and Japanese teams. • Spent three months in factory training, learning the production process while working on multiple lines.

Bachelor Of Engineering (B.E.) With Honours, Mechatronics, Mechanical

Bachelor Of Asian Studies, Asian Politics And International Relations, Japanese

Nanodegree, Machine Learning Engineer

Python For Everyone Specialization