Recently, I lost access to VNC into my raspberry pi. However, thankfully, the Browser streaming (with Picamera) and SSH still work! In order to get my self-driving car tasks to work, I need to appropriate calibrate my line follower and right now I cannot access the Desktop setup or troubleshooting icons. I found the instructions on how to perform the line follower threshold calibration via the terminal - https://www.dexterindustries.com/GoPiGo/gopigo-line-follower-getting-started/using-gopigo-line-follower/. When attempting to run the script, it gets stuck right after ‘Press ENTER’ for setting up the White line [ see attached terminal screenshot ]. I would run the troubleshooting/test to get a log, but I’m unfamiliar with how to do this within the SSH environment.
Please let me know if you need any more information and what steps I need to take next. I look forward to hearing from you soon!
It looks like the line sensor isn’t recognized by the GoPiGo.
As @graykevinb has posted can you run the all_tests.sh with sudo bash /home/pi/Desktop/GoPiGo/Troubleshooting/all_tests.sh, this will generate a log on the Desktop directory, attach the logs with your reply. Also note while running the the troubleshoot program, remember to have the line sensor connected to the GoPiGo.
Another question for you, Do you see LEDs blinking on your Line sensor.
Attached is the log.txt file from the sudo bash /home/pi/Desktop/GoPiGo/Troubleshooting/all_tests.sh command. Please let me know your thoughts and how I need to proceed.
Additionally, I have noticed that the Pi has been running pretty HOT recently during basic connection (no training or image processing). Should I be concerned about this? Or given the size and lack of fan (cooling system), is this typical?
I was finally able to get access to the Remote Desktop/VNC. After getting in, I tried to utilize the Desktop Line Follower Calibrator Tool and for each white and black I was returned (-1, -1, -1, -1, -1). Since I have GUI access now, I went ahead and ran another Troubleshoot/Test script and the output log file is below for your reference.
Maybe this will be helpful in determining the root issue now.
I greatly appreciate your assistance. These things happen and your team has been very helpful and prompt in giving me detailed advisement. So, no worries.
I do need this ASAP though. I have submitted my replacement request to the Feedback section you provided.
I’m a graduate student and this device (along with the GoPiGo) is part of a Deep learning project to simulate self-driving car tasks - initially with a picamera and eventually the ultrasonic sensor. The line follower is, of course, needed to keep this little guy on track so that it can recognize the STOP sign and respond.
We are sending you a line follower sensor ASAP.
Someone will get in contact with you and ask for further details.
Consider it done.
Also, you’ve caught my attention when you’ve said the magic phrase deep learning project.
I’ve also been recently very interested on evolutionary algorithms, so you can understand my excitement.
One idea I’ve been thinking of is to create a design of a car / creature / plane through evolutionary algorithms and afterwards couple what I get with some neuroevolutionary algorithms (neural networks with dynamically-changing topologies).
Even though machine learning is something entirely different from evolutionary algorithms / neural nets, they are quite similar in their philosophy: they both target optimization & learning.