Weekly Report 5

Team Report 5  Integration Plan

This week, Elias and Zhiqi started planning the integration of the suitcase. We need to find the suitcase that is modifiable and cheap(must be within our budget).

Our choice is a Traveler suitcase on Amazon.

Our integration plans are as follows:

• Driving Wheels

Dissemble the original wheels and replace with the Macnum Wheels.

We can dissemble the wheels from inner side of the suitcase without cutting the corner off. But we need to redesign the Bracket due to the size of the Macnum Wheels. Motors could be installed inside the suitcase.

• Following Wheels

We need to mount the wheel on the side of the suitcase without destroying the corner pieces, which would cause the suitcase to shake and become unstable. We will also hollow the middle part of the bottom, and insert a new following wheel to balance. Since there is a seperate layer, the items would not fall out.

• Camera

Camera could be mounted in the front of the suitcase and facing upwards.

• Computing Unit

The whole Jetson board, battery and the IMU could be installed in the suitcase. We need to fix them in a seperate layer.

Zhiqi Report

This week, I almost finished the PID modules, and Elias has almost finished on his part. We decided to move on and started the integration. We have searched for hundreds of suitcase and made the plan based on their price and modifiability.

I have discussed with Elias with specific requirements our suitcase need to have. We went to Amazon and went through hundreds of suitcase. And we discuss each of them to see whether we could disassemble it and instal our own stuffs. More details in the team report installation plan.

Another thing I need to do in the integration is that I need to consider how to connect the Pololu encoded motor to the PID modules. Without this connection, the PID is useless in a sense that we cannot control the speed or read the speed data. Pololu could be driven by arduino. “Actual speed” is computed based upon the encoder informations, acquired in an interrupt routine. By getting the count from the motor, we can calculate the speed as  ((count – lastcount)*(60*(1000/LOOPTIME)))/(16*29);Voltage (“low bat” warning) and current (overheating) are also continuously monitored. Arduino could drive the motor by writing PWM to the system. It will increase PWM in an attempt to maintain actual speed similar to the set speed level by using analogWrite(PWM1, PWM_val);

Elias Report

This week I wrapped around the functionalities for computer vision part, and makes sure that the entire pipeline is working well. Then we started to do the integration together. I did not start the new task on path planning, since we need time to wait for our suitcase to come and build our physical system up.

We are one week behind the schedule, since we just ordered a suitcase. I was testing all computer vision algorithms with the camera set up on my own suitcase. Since the dimensions are pretty similar, I believe when the new suitcase arrives and assembled as our plan, I can just do some sanity checks of the algorithm, and then start to work on path planning part. We will update the schedule and postpone all parts by a one week gap.