Embark
About the company Embark's website
At the beginning of May 2016, I pivoted my previous company, Varden Labs, to develop self-driving technology for commercial big rigs. Varden labs was already a company developing self driving technology, just for a different industry and while working in the space, it became clear to us that there were two major industries that were going to change as soon as self-driving was possible. The first was the space that everyone seemed to be after, taxis. With a clear dollar value of what you were replacing (the driver cost), the economics of expensive first generation self driving technology simply makes sense in the industry. What we didn't understand was that although commercial trucks had a very similar driver that could be replaced, there were very few people focused on that industry, despite its similarly massive market size.
​
With the money that we had raised prior to pivoting, we hired key employees from both the self-driving and aerospace industries and begun developing the first generation prototype, which is now public and can be seen in the video above. At the very end of December 2016 however, I left the company due to continuous disputes with my co-founder of which I cannot publicly talk about. At this time, we had grown the company to 10 people, completed the testing of our initial systems on private land, ready to begin on public roads, and was worth an estimated ~50 million dollars.
​
​
My roll
As president and COO, I obviously had many responsibilities outside of engineering related to the growing and managing the company. This included things such as hiring and figuring out who to hire, managing timelines, managing employees, establishing culture, managing budgets and much more. This process was invaluable and taught me things that I now take for granted and wish I knew before starting. In addition to things related to the company, my roll was heavily focused on engineering. Generally, my responsibility was the development of the hardware systems, all software related to vehicle interfacing and vehicle controls, all mechanical systems such as actuators and sensor mounting and the safety of the whole system. Although I cannot talk about my work in too much detail due to confidentiality, the following outlines chronologically the project that I worked on and some details that have already been made public.
After pivoting, I immediately started focusing on how to control the throttle, brake and steering of an 80,000 lb truck safely as it barrels down the highway at 65 mph. This process began with research into what was already available to suit our needs as well as working with a few very large automotive suppliers to get access to prototype systems as well as information into existing systems. Once we had physical access to our truck, one of our engineers and I worked tirelessly to reverse engineer the CAN networks of our truck using various tools such as Vector CANalizer to get information that we were unable to obtain otherwise. This ultimately led to having full control over the trucks movement and systems with a system that would be considered safe enough by regulators to conduct public road tests with.
​
After we had control over the truck, the same engineer I had been working with began working on designing the hardware systems of the vehicle with my support. During this time, I designed, manufactured and assembled mounts and wire harnesses for all sensors that would be needed for our initial system such as GPS, Lidars, Radars and Cameras among others. The mounts were flexible enough that we could make location changes or sensor changes quite easily. I also worked on interfacing with all sensors as well as synchronizing data acquisition allowing us to begin collecting mass amounts of data to be used for developing our higher level software.
​
Next, I began developing the software that would run on our lower level hardware. Generally, the system was comprised of a higher level system running algorithms pertaining to localization, perception etc which sent information down to a lower level system. I worked on vehicle lateral control from trajectory to steering angle at highway speed, longitudinal control with throttle and brake, driver interfacing and driver takeover detection.
​
Many safety related functions and processes were also performed and were the outcome of an ongoing process of safety analysis that I led. From the start, we designed our systems under the constraints of our functional safety analysis. Together with a number of our engineers, we designed much of how our systems would work using ISO 26262 principles (a functional safety spec for advanced driver-assistance systems). This process also drove the development of our verification and testing practices that we performed on our system. Just prior to leaving, I was also working on hiring a team to begin work on a testing and verification pipeline that would involve SIL/HIL testing on our systems through ongoing development, a practice that we deemed necessary as we began to scale development.
Related Project Links
I was the Engineering Lead of the University of Waterloo's Formula Electric team (FSAE). I lead the engineering efforts of the 2018 vehicle, worked toward establishing a new team culture, and picked up on any project that needed extra hands.
I tested robot localization algorithms in ROS (robot operating system) including particle filter based localization using a Kalman filter on state change and LiDAR based scan matching using ICP for prediction and the ICP fitness score for resampling
As part of a team of 4, we are developing an underwater vehicle to travers an obstacle course in the fastest time possible. My primary roll is to develop the GNC (guidance, navigation and controls) stack running on an onboard android phone
When learning more about modern control theory, I wanted to try a more challenging project that I had wanted to design for a while. Using Simulink, I designed a simulation along with an LQR controller to balance a double pendulum even when subjected to noise.