Final reports and code submission are due Sunday 11:59pm.
These instructions apply to ALL projects.
Every student must submit a PDF file with your individual report through WebCMS or with the following give command:
give cs3431 Report2 <your file>.pdf
Each group must submit their source code. Only one member of your group should submit the ROS source directory in compressed TAR or ZIP format (maximum size 100 MB). Include a README file that explaini how to run your code.
Only submit the src directory. DO NOT submit the whole workspace.
Submit your code with the give command:
give cs3431 Code src.tar.gz
or
give cs3431 Code src.tgz
or
give cs3431 Code src.zip
You can view your COMP3431/9434 marks by logging in to sturec:
https://cgi.cse.unsw.edu.au/~give/code/login.php?app=/~give/Student/sturec.php
As last week, please use the labs to keep working on the second stage of the project. Remember that demonstrations are due in your week 10 lab session and the final report is due at the end of week 11.
There are not lectures. You can use that time to work on the project. If you plan to come into the lab please let me know.
The default resolution of the TurtleBot3 camera (actually a Pi Camera) is 640 x 480 with a refresh rate of 10 frames a second. This is too high for the network and you don't need this resolution to recognise the coloured beacons. You can reduce the resolution to 160 x 120 at 2 fps and still find the beacons. This greatly lowers network traffic.
Week 7, 8 and 9 labs will be for you to continue working on the second part of the project. The tutors will be there to help, but this is your time to work with the robots.
TurtleBot groups will also be able to use the lecture time 4-6pm Tuesday, though there will be minimal supervision during that time.
Other project groups, please consult your tutor.
We'll aim to do the demos for all groups in week 10, with reports due at the end of week 11.
Several students have had their VMware trial licence expire. This would happen if you downloaded the full version of Workstation on Windows or Fusion on macOS. If you download just the player version, this should have no licence limitation and you can still run the class VM.
I've posted the specifications for the second part of the TurtleBot project. If you are doing one of the other projects, please talk to your tutor.
There will be no lectures in week 6, and from week 7 onwards, we will use the lecture time as additional lab time for anyone who want to come in, although there will be minimal supervision.
We still have lab exercises in week 6
. These will be starting tutorials on transforms and computer vision that you will need for the next stage of the project. You won't need the TurtleBots for most of these tutorials, so you can do them in your own time, but tutors will be on hand during the lab times.
Give
has been enabled so you can upload your reports. You can submit your PDF file via WebCMS or using the command line on a CSE machine:
give cs3431 Report1 <your file>.pdf
You are only required to upload the report.
Some students have been having problems seeing the TurtleBot topics on their Windows machines. They can ssh to the TurtleBot but cannot list the topics. Apparently there is a known problem with Windows 10 and certain WiFi chip sets that blocks multicast, which is needed for ROS2.
The only remedy that I have been able to find, searching the web, is to go to the device manager, disable and re-enable the network interface. Not exactly an elegant solution, but if you are having this problem, please try it and let me know if it works.
Please note that the due date for the report on stage 1 of the project is 11:59pm Sunday .
This week you will be demonstrating the first stage of your project. Groups doing the special projects (@Home, Pose and Gaze, Kinova Arm and Nao robots) follow the instructions of your tutor. At the end of the week, all students will submit
individual
reports via the
give
system (not yet enabled).
The report should contain the following:
For the TurtleBot groups, run cartographer during your demo:
ros2 launch turtlebot3_cartographer cartographer.launch.py
Make sure you enable trajectory visualisation in Rviz and, at the end of the demo, take a screen shot of the map, with the trajectory, and paste it into your report.
To help you get started with writing the wall follower code, I've created a src folder with the collision avoidance code from the Gazebo simulator example, except that I've removed the dependencies on Gazebo. Follow the instructions here to download, compile and run the code.
This should get the robot moving in a straight line and then turn when it is close to an obstacle. You should only have to modify the source file wall_follower.cpp to turn it into a wall follower.
We'll work on this in the week 3 labs but you should try to download and compile it before then. There is a lot of code to set up publishers and subscribers that might be a bit confusing. For now,
don't touch it
. It should work for you without having to make any changes.
The logic to guide the robot around should be fairly easy to understand, and that you will have change.
If you have an M1 Mac, you are probably having trouble getting a VM to work. There may be a solution using UTM. Please try following the
instructions here
. This has worked once, but we can't guarantee it will work everywhere. Thanks to Yi Fu for working this out.
I think most TurtleBot groups have been contacted now, but we will run a lab session on Friday 1-4pm in place of the Thursday lab. If you can't make it, we'll work out something else to make up for the missed lab time.
If you are working on one of the other projects, you should have and alternative time with your supervising tutor.
The lecture recording worked this time and can be found here .
We have created a table of the project groups for COMP3431/9434 , as we think they are. We have had to move some people around to make groups of five. Please check that this is ok and let us know if there's a problem.
If ok, you can create a WebCMS group for yourselves. Please use the default group type. The other types are leftover from previous years
So far, the recording of the first lecture has failed to upload. If this continues, I'll upload another recorded version.
Welcome to COMP3431/9434 Robot Software Architectures. The course outline should now be visible. Please take a look.
Lectures will be online using Microsoft Teams, but all labs will be in-person. See the course outline for locations.
This is the link for the Teams live lectures . They will be recorded and can be downloaded from Teams.
This course is project based and you will be working in our labs, with real robots. We will use the first lab session to start organising you into groups of four or five students. I'll describe the project choices in the first lecture on Tuesday. This year allowing a few different kinds of projects, depending on your interests and how ambitious you are.
This course is an opportunity to work on interesting, practical projects. We will guide you through the process but we also expect you to be able to work independently and use the many online resources to help you. We will go through those in the first lecture and the week 1 labs.
See you there.