Presentation Title

Autonomous 3-D Mapping and Collision Avoidance Using LIDAR and Unmanned Aerial Vehicle

Faculty Mentor

Subodh Bhandari

Start Date

18-11-2017 10:00 AM

End Date

18-11-2017 10:15 AM

Location

9-245

Session

Engineering/CS 2

Type of Presentation

Oral Talk

Subject Area

engineering_computer_science

Abstract

This presentation talks about the work being done at Cal Poly Pomona using an Unmanned Aerial Vehicle (UAV) for autonomous 3-D terrain mapping. A DJI S1000 has been equipped with a Pixhawk autopilot in conjunction with Mission Planner for autonomous waypoint operation. Mission Planner, a flight management software, is programmed to autonomously fly around selected pin-points in a targeted location designed in accordance to control points placed by Civil Engineering students. The sensors used are an XSENS inertial measurement unit (IMU) for position and orientation of the vehicle, and a VLP-16 light detection and ranging (LiDAR) sensor for 3-D Mapping. An Intel NUC is utilized as an on-board flight computer. It will process and store LIDAR and IMU data and implement the collision avoidance algorithm. A point-cloud algorithm will perform geo-referencing calculations using the sensors’ raw data in conjunction with the GPS data to generate a point cloud file. This file is uploaded to LAStools to generate a 3-D visualization. The data will be analyzed and compared using ArcGIS, LAStools, MATLAB, and a survey done by Civil Engineers. For collision avoidance, the LIDAR is used to detect any obstacles in the UAV’s flight path. In the case of any obstacle detection, the flight computer directs the Pixhawk autopilot to re-route the flight path. The UAV then resumes the original flight path once the obstacle is avoided.

Summary of research results to be presented

Through the use of Application Programming Interface (API), we have successfully able to write a program in C++ to extract data from the VPL-16 LIDAR and IMU into a text document. Using various theories of Orbital mechanics, we were able to develop a code that uses NED frame typically use in Aerospace and Navigation applications. This allow us to gather data from a point in respect to where the UAV started flying and with respect to Earth. This code with the integration of the LIDAR raw data creates a point cloud assembly which creates a 3D point map.

The LIDAR data is to be use as our sensor detection for collision avoidance. Studies have taken place to study the control of our UAV, which will help us develop a code that inputs control command for the UAV to move when an object is to be avoided. Also the code must be able to avoid and return to its original flight plan that we set.

A 3D Model has been done in SolidWorks for demonstration purposes and now will be use for animation purposes with LIDAR raw data.

This document is currently not available here.

Share

COinS
 
Nov 18th, 10:00 AM Nov 18th, 10:15 AM

Autonomous 3-D Mapping and Collision Avoidance Using LIDAR and Unmanned Aerial Vehicle

9-245

This presentation talks about the work being done at Cal Poly Pomona using an Unmanned Aerial Vehicle (UAV) for autonomous 3-D terrain mapping. A DJI S1000 has been equipped with a Pixhawk autopilot in conjunction with Mission Planner for autonomous waypoint operation. Mission Planner, a flight management software, is programmed to autonomously fly around selected pin-points in a targeted location designed in accordance to control points placed by Civil Engineering students. The sensors used are an XSENS inertial measurement unit (IMU) for position and orientation of the vehicle, and a VLP-16 light detection and ranging (LiDAR) sensor for 3-D Mapping. An Intel NUC is utilized as an on-board flight computer. It will process and store LIDAR and IMU data and implement the collision avoidance algorithm. A point-cloud algorithm will perform geo-referencing calculations using the sensors’ raw data in conjunction with the GPS data to generate a point cloud file. This file is uploaded to LAStools to generate a 3-D visualization. The data will be analyzed and compared using ArcGIS, LAStools, MATLAB, and a survey done by Civil Engineers. For collision avoidance, the LIDAR is used to detect any obstacles in the UAV’s flight path. In the case of any obstacle detection, the flight computer directs the Pixhawk autopilot to re-route the flight path. The UAV then resumes the original flight path once the obstacle is avoided.