Mehul Aggarwal

Lane Detection Software Project

This semester I am currently working on a group project focusing on the simulation of an autonomous car where we have included lane detection capabilities as well as a closed loop feedback controller system.

We are using the software Gazebo in order to create a virtual environment with our car on a road in order to test our controller algorithms. This virtual environment allows us to create a controlled environment where we can see the progress of the car and how it behaves to the different road conditions we have inserted into the track. To begin with, we used the virtual environment in order to implement our lane detection algorithm. Our algorithm involves a combination of gradient filtering and colour filtering in order to create an image of road which is better interpretable by the computer. After we were successful in making our lane detection algorithm in the virtual environment, we applied it to footage of a real car moving in different real road conditions. This allowed us to fine tune our parameters so that they worked in these real conditions.

Once we knew that our lane detection perception algorithm was working, we moved on to our car controller system. We implemented a PID controller for the car. This involved creating a state model for the car which included a closed loop controller which changed the velocity and orientation of the car based on the current position, velocity and orientation. We used our Gazebo virtual environment in order to tune and test the PID controller parameters. We created a waypoint system in the simulation which outlined the track in the state model parameters for the car controller. This allowed us to test our feedback system on a track with complex turns and curves.

Finally, we combined our lane detection and controller algorithms together so that the car was just moving based on the lane information we were interpreting form the car camera. We did run into a few problems here though due to the perception algorithm creating a track which the controller was not able to follow. This was solved by tuning the controller parameters to allow more aggressive movements and tuning the lane detection algorithm parameters to create a road image with less noise which is more easily interpretable.

Our next goal is to implement our algorithms on a testing hardware car to analyse how the algorithms work in non-ideal real-life conditions.

What I Have Learned From Working in a Team

This semester I am currently working on a group project focusing on the simulation of an autonomous car. This is a software project with a group size of 5 people, including me. It has been an extremely interesting experience working on a coding project with such a large group and here are 3 big things I have learned about working in a team like this:

1. Ensure everyone in the team shares their ideas. This has been especially important because we are a relatively large group. There have been many times while we have been working on the project when different people have had different ideas on how to implement a particular algorithm. By discussing everyone’s ideas, we were all able to get a better understanding of the algorithm because we could see the different ways in which it could be implemented. In addition, after a good discussion, we were able to combine different ideas together to create an efficient algorithm that everyone was happy with. This wouldn’t have been possible without the free flow of ideas.
2. Utilise everyone’s strengths. When we first got together as a group, we discussed what each team member’s strengths were. By doing this, we were able to easily divide up the work that we needed to complete and efficiently complete the project because everyone was working on what they were good at. Furthermore, after completing certain tasks, we wanted to make sure that everyone understood how everything worked, and so each person would explain how they completed their assigned task and how their part works. This allowed us to effectively complete the project but also develop everyone’s skills at the same time.
3. Communication is essential. It is highly inefficient for different team members to be working on the same task or people not talking about when to schedule meetings or how to combine different sections of code together. We made sure to have regular meetings with each other to work on the project and assign tasks, and constantly communicated progress on the group chat to keep everyone on the same page. Communication like this is extremely important to make sure everyone knows what they are doing and to keep everyone involved. If there is a culture of open communication, everyone will feel more comfortable to share ideas and ask for help and thus work to the best of their ability.

Experiences of Living in Australia and USA

I have grown up in Australia since the age of 7 and decided to come to the US for university. I have now spent more than 3 years at university in the US and discovered some interesting differences between living in Australia and USA.

There is a big difference in the work-life balance and lifestyle between the 2 countries. Australia in general has a much more laid-back and patient lifestyle, whereas USA has a much more fast paced and “time is money” lifestyle and attitude. This is probably due to the enormous size of the US, which means there is a lot more competition, and a much higher need to work harder and prove yourself in order to get opportunities. Even in university, I know from some my friends in Australia, that they usually won’t have classes for more than 3 days a week and most people have a part time job while studying at university. At university in the US however, I have classes for 5 days a week and not many people have part time jobs because there is so much class work, homework, projects, as well as extracurricular activities such as participation in student organisations and research projects.

Moreover, the sense of humour in the 2 countries can sometimes be quite different. On important aspect of a country’s culture is its sense of humour. A comparison between American humour and British humour would tell you that it’s a battle between American optimism and British cynicism. Australian humour is a hybrid of the 2 – firmly on the British side, but with a signature air of ‘no worries’ that mirrors American optimism. I sometimes find that I might make jokes which don’t really translate to my American friends but would have been funny to my Australian friends. However, overall, this isn’t a big impedance in making friends, but more of an interesting thing to mention that the 2 different cultures have different styles of humour.