Self-Driving Engineer

Training Base

Playground For Engineers

Self-driving Engineer

Training Base

End of 2019

Pre-Signup

Course Contents

Practice and test of Autoware algorithms on real cars

General course about Drive-by-wire and Car Electronics

Workshop:

Metal 3D printing in car manufacturing

Training and lecture of Autoware, the world's first open source autonomous driving software

Autonomous driving road tests

Complete challenges from Udacity course such as traffic light recognition, lane detection etc

Hardware Support

5 self-driving development cars
Full set of sensors – LiDAR, industrial camera, GPS inertial navigation, millimeter wave radar, etc.
Various industrial CNC machining equipment
Metal 3D printing system
Training venue, testing roads and setup , application scenes
By-wire control, vehicle retrofitting, racing events
Real car Hands-on practice creating Autonomous Driving Systems with the world's first open-source software“Autoware”

The Training Site

PIX FACTORY

5000 square meters
Test site in industrial park

The Open Road Test Site

URBAN AREA

One square kilometer's urban road In Guiyang Hi-tech district

GUIYANG,

CHINA

Mentors

Alexander Carballo Segura

Autoware Software Mentor

Designated Assistant Professor at Nagoya University, Research Consultant at TierIV Inc. Doctor of Engineering (D.Eng.) in Computer Science from University of Tsukuba, Japan. Member of the Institute of Electrical and Electronics Engineers (IEEE) Robotics and Automation Society (RAS), Computer Society, and Systems, Man and Cybernetics Society. Professional Member of the Robotics Society of Japan (RSJ). Active reviewer for IEEE Int ref Intelligent Robots and Systems (IROS), also Advanced Robotics Journal and the Journal of the Robotics Society of Japan. Great research and work experiences in the autonomous driving industry. Main research interests: Human-robot interaction, social robotics, autonomous robot navigation, autonomous vehicles, robotic perception, machine learning, computer networks, unmanned aerial vehicles.

Dowson Tseng

Vehicle Hardware Mentor

Full-stack engineer at PIX. Deep proficiency in car hardware and software. Specialized in applying software and controls into vehicles to realize autonomous driving applications in real scenarios. Majored in Automation, Dowson holds years of experience in autonomous driving and have participated in the mechanical designs of PIX Robo-vehicle and coffee car. Dowson is the main lead in the mechanical and drive-by-wire controls of PIX cars, with rich engineering experience in the autonomous driving upper system.

Agenda

DAY 1

9:00 AM - 12:00 PM

• Comprehensive introduction: Self-driving Training Base and PIX self-driving car
• Overview: Onsite training tasks, goals and daily training contents
• Introduction: The world's first self-driving open source software - Autoware
• Warming-up: Catch up and get to know each other; 5 students form a team and choose their own car

• Safety Training Session

13:30 PM - 18:00 PM

• System and software deployment of each team
- Tensorflow installation (GPU version)
- Autoware installation
• Day1 summary, wrap-up and sharing

DAY 3

9:00 AM - 12:00 PM

• Team practice on physical car
- Connecting sensors, opening software, collecting and mapping, and enabling the basic self-driving mobility in each team’s car

13:30 PM - 18:00 PM

• Team practice on physical car, modify the Autoware source code and add the project to Autoware framework, enabling the traffic light recognition function of the course in Autoware
• Day3 summary, wrap-up and sharing

DAY 5

9:00 AM - 12:00 PM

• WORKSHOP:
- Courses of By-wire Control, Electric Vehicle Generative Design and Metal 3D Printing
- Keep optimizing the task performances

13:30 PM - 18:00 PM

• Group JAM: Bring out each team's creativity to realize more functions on the self-driving car under the premise of safety.
• Show & Tell from each team

DAY 2

9:00 AM - 12:00 PM

• WORKSHOP:
- Applying Autoware On Cars, Professor Alex Segura from Nagoya University
- Further understand Autoware framework and the relationship between each node, practice and learn map collection

13:30 PM - 18:00 PM

• Test onsite for data collection and training
- 1. Data collection, labeling, training targeted detector
- 2. Configure the relevant files
• Test the classifiers and get the classifier output
• Day2 summary, wrap-up and sharing

DAY 4

9:00 AM - 12:00 PM

• Team practice on physical car
- Test the traffic lights and check the Start/Stop functions

13:30 PM - 18:00 PM

• Operation on-board: Optimize the detection and car functions
• Day4 summary, wrap-up and sharing

Why

Engineers need to practice and test self-driving technology on the physical cars, learning both software and hardware comprehensively

The global demand for self-driving talents is about 100,000+, but existing full-stack engineers number can only meet 10% of that demand;
Running the algorithms on the physical hardware and vehicles will help engineers improve their development capabilities;
Trillion market demand, 10,000+ startups, 1000+ traditional car enterprises, are all looking for talented self-driving engineers, whose number is only 10,000;
Self-driving development is highly dependent on testing and data, while there are only 30+ self-driving test sites around the world, with strict requirements;

They Said

“I think (establishing the training base) this is a very good idea. After engineers finish their courses, they’re gonna realize all those on the real car.“

Zepeng Lee

"The hardware here is very complete. It's a great honor to be among the first batch of engineers to attend the training. I hope the best for PIX Self-driving Training Base."

Hao Jiang

I am so honored to participate in the first cohort of the self-driving training base. During the five days of training, I have gained a lot. I believe that more students will come here to learn, and the training base will do better, which is obvious. The training base will provide more for the students.

Engineer Experience

& Feedback

”

"I think (establishing the training base) it’s a great concept. Whether it’s company employees or hobbyists, they all lack the experience from online coding to onsite practice, which is quite necessary for a self-driving engineer."

Si jun Xiao

I worked together with my teammates from coding to software application. Finally, we completed the whole process of the hardware configuration and vehicle debugging on real car. The five-day experience is quite impressive to me and I’ve learned a lot.

Qinghua Yang

Shenting Yu

Angelo Yu

Founder & CEO

of PIX Self-Driving

When we started PIX project, the biggest problem we encountered was the lack of engineers. Many engineers enter the autonomous driving industry by changing careers from other related fields such as robotics, automation, image search, artificial intelligence, and so on. One of the biggest problems after doing that is the engineers know little about the vehicle control and the hardware, and they don't have the experience of running algorithms on the real vehicles. So we hope to make up to the lack of onsite training through the training base.

So far in the world, there are few such places. The thing is that universities teach researchers or teach engineers in very specific fields like robotics only or vehicle control only, but being able to integrate all the technology and being able to deploy that into a car, it’s just so a few places. The most important part is that being able to do that on the real car as PIX is proposing is an amazing idea so I think it’s going to have a big impact not only in China, but it will attract the attention of the world and people will be applying from the world to come here. So I think that there is transformation that PIX’s unique idea of putting people in the practical scenes, putting people in the cars and being able to actual control the cars.