Month: April 2019

Name: Wenxin Xiao

• What did you personally accomplish this week on the project? Give files or photos that demonstrate your progress. Prove to the reader that you put sufficient effort into the project over the course of the week.

• • Over this week, I worked to experiment on the temperature and angle threshold that would work best with the lidar. The thermal camera can now communicate with the lidar to obtain distance information on the obstacle observed and return to lidar when it flags a heat signature with help from Aditya and Akshat.

• Is your progress on schedule or behind? If you are behind, what actions will be taken to catch up to the project schedule?

• • Our progress is slightly behind. We will work extra hours to catch up to the project and ensure we have the integration of the project and record many testing scenarios.

• What deliverables do you hope to complete in the next week?
• We aim to test on the thermal part of the project to ensure that it is identifying humans correctly and marking on the correct locations on gmap. We will mostly focus on testing the navigation, mapping, and human identification on multiple scenarios multiple times.

Name: Akshat Jain

• What did you personally accomplish this week on the project? Give files or photos that demonstrate your progress. Prove to the reader that you put sufficient effort into the project over the course of the week.
• This week, I mostly helped Wenxin and Aditya set up the thermal camera so it can corresponding distance information from the Lidar as soon as the thermal camera flags a person. I also worked with Aditya to fix a bug we had our robot when it would try to move back while executing a trajectory. We tested our robot in Hammerschlag to fix this problem.

• Is your progress on schedule or behind? If you are behind, what actions will be taken to catch up to the project schedule?
  • Our progress is behind as we were hoping to get the path planning working in time for the demo. We will mostly be working on this to fix the bug.

• What deliverables do you hope to complete in the next week?
  • We aim to fix this bug by the begining of next week. We also hope to extensively test the thermal camera and its ability to place a human on the map and its accuracy in doing so.

Aditya Ranade

• What did you personally accomplish this week on the project? Give files or photos that demonstrate your progress. Prove to the reader that you put sufficient effort into the project over the course of the week.

This week, we were primarily focused on getting the thermal camera set up for the final demo. I specifically worked on getting range data from the lidar offset whenever the thermal camera detects a person. I also worked on testing frontier exploration in the HH hallway and there appears to be an issue regarding execution of the planned path. The robot sometimes did not follow the path and moved backwards instead. I will work on fixing this issue in the first half of next week for the final demo.

• Is your progress on schedule or behind? If you are behind, what actions will be taken to catch up to the project schedule?
  • I would have liked to have frontier exploration working decently well; however, the issue mentioned above still persists. I don’t have much other work this week, so I will work on fixing this bug at the soonest.

• What deliverables do you hope to complete in the next week?
  • Fix the bug in frontier exploration
  • Integrate thermal camera with exploration.

TEAM UPDATE

This week, we got made progress on the human detection subsystem. We attached the thermal camera to our mount and ensured that it is flush with the lidar 0th degree. We determined thresholds for human detection and the thermal camera is able to detect people within a range of 1.5-2m. Overall, we are behind schedule since we did not anticipate this bug which is causing exploration to fail at times. We plan on devoting more time towards our project this week to fix the bug and fully integrate the thermal camera into our exploration.

Name : Akshat Jain

This week I worked on 3D printing the stand for the Raspberry Pi, Power Bank Thermal Camera as well as the Lidar sensor. We assembled the 3D print and attached all of the parts to the iRobot. In addition to this, we worked on creating an appropriate transformation for the iRobot along with our stand so that it works in simulation. We also started initial real world testing in basic rooms. One problem we are facing is that the Local network created by our laptops is not powerful enough to communicate with the robot. We plan to use a router to create a much more powerful local network to fix this problem.

Name: Wenxin Xiao

Team Members: Aditya Ranade, Akshat Jain

• What did you personally accomplish this week on the project? Give files or photos that demonstrate your progress. Prove to the reader that you put sufficient effort into the project over the course of the week.

• • Over this week, I worked on the code for the nodes that allow thermal to communicate with lidar to obtain distance information and publish the location of detected objects. The frame and the legs to uphold lidar got 3D printed. The team mostly focused on connecting to hardware components and move from simulation to actual mapping.

• Is your progress on schedule or behind? If you are behind, what actions will be taken to catch up to the project schedule?

• • Our progress is slightly behind. We will work extra hours to catch up to the project and ensure we have deliverables by demo next week.

• What deliverables do you hope to complete in the next week?
• Next week we will mostly work to test our autonomous mapping and obstacle avoidance with our test scenarios and then move on and move on to thermal. Since a frame is not given in the specification, I will perform real tests to approximate the frame in order to obtain the angle of the object detected and connect thermal to other parts. We will then move on to test thermal detection with later test scenarios.

 

Name: Aditya Ranade

• What did you personally accomplish this week on the project? Give files or photos that demonstrate your progress. Prove to the reader that you put sufficient effort into the project over the course of the week.
  • I worked on testing frontier exploration in hardware and real-world environments. At first, the code was running really slowly, so we made some optimizations regarding the map size and added threads to make some actions concurrent. We saw decent results as the robot was able to navigate to unknown spaces. We will continue working on tweaking our code to improve results.
• Is your progress on schedule or behind? If you are behind, what actions will be taken to catch up to the project schedule?
  • Our progress is on schedule. We will be working on human detection using the thermal camera this week.
• What deliverables do you hope to complete in the next week?
  • We will test the frontier exploration subsystem in more real-world settings. I will work on improving the speed and correctness of this subsystem.

Team Report

This week we 3D printed our frame to hold the lidar and other parts on the iRobot. We worked to assemble the robot and run frontier exploration and gmapping on it.

Next week we will start real world testing with our frontier exploration and mapping with various test cases. We will also mostly focus on the thermal part of the project, debugging and modifying the algorithm to ensure it works in real life, and start real world testing with our thermal detection with various test cases.

Name: Wenxin Xiao

• What did you personally accomplish this week on the project? Give files or photos that demonstrate your progress. Prove to the reader that you put sufficient effort into the project over the course of the week.

• • Over this week, I modified the code for human detection and investigated methods of creating a node for thermal to communicate with other parts of the design as no available package was found on the thermal camera we purchased.

• Is your progress on schedule or behind? If you are behind, what actions will be taken to catch up to the project schedule?

• • Our progress is on schedule. We have each part working separately in simulation and will start integrating the components together.

• What deliverables do you hope to complete in the next week?
• Next week we will mostly work together to try to move from simulation to the real robot chassis. The board should be 3D printed and placed on the robot and so should all other parts. The frontier exploration and thermal camera need to move from simulation to real world testing.

 

Name: Aditya Ranade

• What did you personally accomplish this week on the project? Give files or photos that demonstrate your progress. Prove to the reader that you put sufficient effort into the project over the course of the week.

• • Over this week, I fixed the obstacle avoidance code, which was an issue in the mid-semester demo. I increased the obstacle detection threshold and also increased the size of obstacles by changing the values of cells adjacent to obstacles to being an obstacle itself. The preliminary results have so far been good as the robot does not collide into walls or other obstacles.

• Is your progress on schedule or behind? If you are behind, what actions will be taken to catch up to the project schedule?

• • We are on schedule for the frontier exploration and hardware implementation. We need to catch up on human detection. We will be spending most of the next two weeks on this problem.

• What deliverables do you hope to complete in the next week?
  • We will print the 3D mount for the lidar, thermal camera, RPi and battery pack. We will assemble the robot and test our code in real world scenarios.

Name: Akshat Jain

• What did you personally accomplish this week on the project? Give files or photos that demonstrate your progress. Prove to the reader that you put sufficient effort into the project over the course of the week.

• • Over this week, I fixed some issues we were facing during the mid-semester demo with the simulation. I updated the lowest cost path heuristic to add a cost to all cells on the grid adjacent to obstacles. Our robot now turned farther away from the obstacles than it initially was. I also set up the LIDAR with ROS so that it could start sending point cloud data to our frontier exploration subsystem.

• Is your progress on schedule or behind? If you are behind, what actions will be taken to catch up to the project schedule?

• • We are mostly on schedule for individual subsystems. We will be combining all of them together in the upcoming weeks.

• What deliverables do you hope to complete in the next week?
  • We hope to have the 3D mount for all the hardware components printed so we can start  real world testing.

Team Report:

Overall, we have made very good progress on autonomous exploration using the heuristic of finding frontiers and moving to the cluster with the greatest mass in order to increase the amount of new information we get about our environment. The next step is to test this on the iRobot Create using the Lidar sensor. We expect no significant issues to arise when testing this in hardware.

On the human detection front, we are slightly behind since we require the robot to be setup. As a result, the first priority is 3d printing the mount and assembling the robot to allow for quick real-world experimentation.

Name: Wenxin Xiao

Team Members: Aditya Ranade, Akshat Jain

• What did you personally accomplish this week on the project? Give files or photos that demonstrate your progress. Prove to the reader that you put sufficient effort into the project over the course of the week.

• • Over this week, I mostly focused on moving from simulated data to actual testing of the code for human identification. I worked on the wiring and connection between raspberry pi and thermal camera, configured the raspberry pi to allow my code to be run, and initialized i2c connection with the sensor with Akshat. The thermal camera now outputs 8×8 thermal information correctly and I further interpolated it to allow for 32×32 thermal information.

• Is your progress on schedule or behind? If you are behind, what actions will be taken to catch up to the project schedule?

• • Our progress is on schedule. We have a working frontier exploration, a working obstacle avoidance, and a working thermal image output that identified heat signature separately as planned for the demo.

• What deliverables do you hope to complete in the next week?
• I plan to modify the code to reduce false positives and false negatives since the working thermal output indicate that human thermal signatures are not in high thirties degree celsius but rather low twenties, close to no human presence. I further plan to write the publisher and subscriber node to connect thermal cameral outputs to the ros network.

 

https://github.com/adafruit/circuitpython

https://learn.adafruit.com/adafruits-raspberry-pi-lesson-4-gpio-setup/configuring-i2c

https://www.mouser.com/datasheet/2/737/adafruit-amg8833-8×8-thermal-camera-sensor-1396559.pdf

Name : Akshat Jain

This week we received most of the hardware components so we started work on setting up each one of them. I set up the LIDAR and connected it to the Raspberry Pi with a serial cable. I was then able receive point cloud data from it. One issue we had was that the point cloud data from the LIDAR did not give one point per degree which it claimed to give originally. We are currently trying to find a fix to this problem.

This week we finally received the iRobot Create 2 as well. I set up the libraries on the Raspberry Pi as well as the lab computer for robot control. We can now give commands to the robot to move with a specific speed and turn specific angles.

I also helped Wenxin set up the Thermal camera and set up libraries for it on the Raspberry Pi. We were able to receive an 8×8 grid of temperature values which would successfully give us a higher reading when we put our hand in front of it.