A Pet Companion Robot

WEEK 5 Date: 28 February 2020 Place: EE Building, 3th floor/ Room 304, 2P22 Before lab l   We were working on code writing of the automatic part this week. l   During this weekend, we discussed how to let the robot react when the recognition system detects the pet. We thought up an idea to determine the distance between pet and robot. When our pet is close enough to the robot, the robot will move to the opposite way. l   Team members spent a number of time on writing the program. During lab l   Test the manual control function and it goes well. l   Try to run the automatic part but fail. Problem and solution l   we met problems when we try to test the code of auto-control function. When we launched the Python interpreter on the command line and then input: import Tensorflow as TF It showed: ImportError:  No module named Tensorflow all the time. The reasons may be one of Tensorflow's dependen...

WEEK 4 Date: 21 February 2020 Place: EE Building, 3th floor/ Room 304, 2P22 Before lab l   Working on installing Opencv and Tensorflow on our Raspberry pi. l   Learning how to get the data from the object detector in Python. During lab l   Run the Tensorflow’s recognition system successfully. Now we can identify our pet through the camera on robot. The number of frames is only 5 frames, but this is the highest frames we can get. After that, we combined it with the motor-controlling program so that our robot can respond when it detects our pet. l   We were struggling with motor issues the whole afternoon and solve them at the end of this lab. Problem and solution l   The main problem we faced today is the motor did not work. When we ran the Python program, it was expected to show a real time image and had four buttons to control the movement of our robot. However, when we tried to push the button, there were some feedback from the R...

WEEK 3 Date: 14 February 2020 Place: EE Building, 3th floor/ Room 304, 2P22 Before lab l   Finish installing mjpg-streamer which is for the remote monitoring module and finish constructing the webpage. l   Learning the way to achieve remote controlling function based on bottle. l   After completing automatic mode contemplated portion, we would use Tensorflow lite to achieve the object recognition function, which is based on Opencv. The program language we used was python3. During lab l   We tested the monitoring function based on mjpg-streamer and it worked well. We also improved quality of images and fluency, and our web design is simple and pretty. l   Start working on automatic part. The first step is to build the recognition system. We would use Tensorflow lite to achieve this function. TF lite is an open source library for object recognition, which is based on Opencva — library of programming functions that can shows real-ti...

WEEK 2 Date: 7 February 2020 Place: EE Building, 3th floor/ Room 304, 2P22 Before lab l   All of us were learning how to use raspberry system since none of us used it before. l   We started learning Python3 since we would use it to write the automatic part’s program. During lab l   Achieve SSH of pi via Wi-Fi and change remote computing toolbox from Putty to Mobaxterm. l   Download all the files we need into raspberry pi. l   Try to use Arduino and Raspberry pi together but decide to use pi to control all of the function finally. l   Start working on the manual control part, and the first function we want to achieve is remote motoring. Problem and solution In the beginning, we wanted to implement web control using Webiopi — a web framework that can control Raspberry Pi GPIO. It did not work and we found out the reason is that the Webiopi official website was no longer updated, and the latest version of...

WEEK 1 Date: 31 Jan 2020 Place: EE Building, 3th floor/ Room 304, 2P22 Supervisor: Prof. James Walsh Group member :  Chunyan Xue Lixiang Yuan Congyi Xu During lab l   Two group members assembled the robot while another one set up raspberry pi. l   We wrote the Raspbian system into TF card and used Putty to SSH our pi, the Raspbian desktop appeared successfully. We also tested the pi camera such as taking pictures and recording videos. Problem and solution l   When we assembled the robot, we broke the wire of one of the motors accidently. This accident is solved by borrowed a soldering iron and soldered it. We didn’t test the function of our robot yet since we forgot to order batteries. l   We met a lot of problems when we used raspberry pi at the first time. In the beginning, we SSH pi via network cable, and we had to find the pi’s IP address from command prompt since it changed its terminal every time we restart it....