We are a San Francisco based team building self-driving semi trucks. We have raised $117MM in total and are backed by Tiger Global Management ($70MM Series C) and Sequoia Capital ($30MM Series B). We move freight daily between LA and Phoenix using our purpose built transfer hubs. This is an incredibly exciting time for autonomous driving and our team is looking to grow.

We are looking for an engineer to help develop and maintain software and processes for calibrating sensors on our self-driving trucks.

This role would be part of the Perception Team. The Perception Team is responsible for designing algorithms and applying machine learning to create in-truck software, for the purpose of creating a reliable statistical three-dimensional model of the relevant environment around our self-driving trucks, using all available sensor, vehicle, and map data. This resulting World Model is used by downstream in-truck software components to plan and execute truck behavior.

In order to accurately and reliably use truck sensor data, we need to accurately know internal sensor parameters, as well as the location and orientation of the sensors with respect their host truck. These critical values change across different individual sensors, trucks, operating conditions, and time. It is the job of calibration software and processes to continuously estimate these parameters while taking into account all their relevant variations. Calibration engineers develop and maintain the software and processes, but also monitor calibration operation in the field to derive continuous improvements, and respond to calibration concerns.

Responsibilities

• Designing and implementing methods for characterizing sensors and truck dynamics
• Analyzing sensor outputs to identify trends
• Designing and implementing methods for estimating internal sensor parameters, as well as the location and orientation of the sensors with respect their host truck, both for off-line and in-truck execution, and both for static and operating trucks
• Selecting or building, and maintaining and operating, software, processes, and instrumentation, for establishing accuracy and reliability of the developed calibration methodologies
• Building and maintaining software tools to monitor calibration performance in the fleet and identify deficiencies
• Diagnosing calibration concerns

Experience Should Include:

• Experience with C++ and Python
• Knowledge of camera multi-view geometry, and in particular the use of extrinsic and intrinsic calibration matrices
• Hands on experience with robotic systems

Experience Might Include:

• Experience similar to the responsibilities of the role
• Familiarity with the state-of-the art solutions to calibration problems 
• Experience with state estimation, parameter estimation, and in particular approaches related to Simultaneous Localization and Mapping (SLAM)
• Experience with cost optimization techniques (numerically minimizing linear and non-linear cost functions, including gradient descent-type and bundle adjustment-type techniques)
• Understanding of operational principles of lidar, radar, GNSS (GPS), and inertial sensors
• Experience developing real-time software
• Exposure to, or demonstrated interest in, autonomous vehicles development
• Education in engineering, computer science, math, or physical sciences
• Ability to communicate concisely but unambiguously in written and spoken form

When you apply, address the application to Michael and let me know why you want to join our team.

A few company highlights:

Embark Blog - Series C and Transfer Hubs

Forbes - 70 Million Dollar Series C

Video - Day in the life of a self-driving truck

Embark Blog - Disengagement Report

30 Million Dollar Series B led by Sequoia