Download as docx, pdf, or txt
Download as docx, pdf, or txt
You are on page 1of 3

UAB ME/EE Senior Design Topic Proposal Form Fall 2020 – Spring 2021

DESIGN AND IMPLIMENTATION OF VISUAL


SENSORING SYSTEM FOR SMALL
AUTONOMOUS GROUND VEHICLE
Mentor’s names Company name
Vladimir Vantsevich
Southern Company Services, Inc.
Siyuan Zhang
Mentors’ email Mentors’ phone #
vantsevi@uab.edu
205-975-5855
zsy1010@uab.edu
One of the major tasks of designing an autonomous driving system is to make autonomous ground
vehicles (AGV) to perceive its surrounding environment. To improve the environment perception
ability of AGV, visual sensoring systems should be designed and implemented. In this visual
sensoring system, a 3D LiDAR and a digital camera can be employed. The 3D LiDAR is utilized to
detect the objects surrounding the vehicle while the digital camera is utilized to capture the images
of the vehicle front view for recognizing the road obstacles, which cannot be detected by the LiDAR
system, such as puddle, grass, gravel, etc. By processing the detected information and captured
images, the distance from road obstacles and the objects to the vehicle can be determined, which
lays an essential importance for the autonomous driving system to plan a collision free and an
optimal trajectory path.

Our research team is seeking a team of motivated senior design students to design a visual sensoring
system for a small AGV, which allows for representing road obstacles and objects surrounding the
vehicle and determining their distances to the vehicle. The students are required to:

1. Design and build a visual sensoring system for a small autonomous ground vehicle shown
in Figure 1.

1) Conduct literature search on 3D LiDAR and digital camera and select suitable ones, which can
be connected and read by the embedded PC employed on the small AGV.

2) Design the layout of the LiDAR and camera on the AGV, which allows the two sensors to have
an open-view and optimal viewing angle to the environment.

3) Design a block diagram in the software RTMaps. With this block diagram, the information
about 3D map of the vehicle’s surrounding objects (detected by LiDAR) and front-view-image
(captured by camera) can be received and viewed by the embedded PC.

4) Design a data processing algorithm in Simulink for data post-processing. This processing
algorithm should be able to read the 3D map data and the front-view-image generated by RTMaps.
Then, by using the collected 3D map data, this algorithm allows to reshape the objects surrounding
the vehicle and represent them as unions or intersections of basic solid primitives (cuboids,

ME Senior Project 2020 - 2021_Vantsevich, Zhang


UAB ME/EE Senior Design Topic Proposal Form Fall 2020 – Spring 2021
spheres, cylinders, and cones). Ultimately, the distance between the vehicle surface and the
reshaped object surface will be determined via the algorithm.

Moreover, this processing algorithm should be able to reduce the resolution of the front-view-
image. A color filter will be designed to filter the color of road obstacles (e.g. puddle, grass, gravel,
etc.) which cannot be detected by the 3D LiDAR. Based on the camera’s viewing angle, the distance
between the vehicle and these obstacles will be determined via the algorithm.

2. Evaluate the performance of this visual sensoring system by testing AGV in off-road
terrain conditions

The 3D LiDAR and camera, and all computer programs designed in RTMaps and Simulink will be
connected and uploaded to the dSPACE embedded PC (Figure 2) installed on the AGV. An
experimental test will be conducted to evaluate the performance of this visual sensoring system.

3. (Optional) Based on the determined distance (from vehicle to objects and road obstacles)
from item 1, design a path planning strategy for designing a collision free and an optimal
trajectory path for the small AGV.

SKILLS REQUIRED

Discipline/Student Percent involvement Specifics


s
1) Layout and mechanical design of the two systems
ME
65% for AGV
(2) Students
2) Integration of the two systems with AGV
1) developing data receiving computer code in
ECE Simulink for embedded PC
35%
(1) Student 2) post processing of sensor data with the use of
RTMaps and Simulink

LEARNING OBJECTIVES

 Completing the entire project would help the students to have a better understanding of the
process of conducting a research, from proposing to demonstrating the theory. They will
also learn how to collaborate with other project members, task management, etc.
 First-hand experience of designing the mechatronic system used for the autonomous
driving systems for electric vehicles.

ME Senior Project 2020 - 2021_Vantsevich, Zhang


UAB ME/EE Senior Design Topic Proposal Form Fall 2020 – Spring 2021
o Process of design of experiments to evaluate performance of a real-world system
with single variable.
 Understanding of data acquisition system (DAQ) for unmanned ground vehicle
o Familiarity with Matlab Simulink and RTMaps implementations
o Use of industry-scale hardware for real-time DAQ implementation

TECHNICAL SUPPORT

Dr. Vantsevich and his student researchers at the Vehicle and Robotics Engineering Lab will work to
facilitate the efforts of the senior design project team in all aspects of this project.

 Dr. Vantsevich/ Siyuan Zhang will meet with the team weekly to review their progress and
provide feedback/direction as required.
 Dr. Vantsevich/Siyuan Zhang will delegate additional student researcher support as needed
to assist team with:
o Access to 3D LiDAR and digital camera
o Access to unmanned ground vehicle
o Access to dSPACE hardware and other electronics required

FINANCIAL SUPPORT

Dr. Vantsevich has access to a number of financial resources that can be brought to bear to fund
procurement of needed materials and fabrication services in support of the team’s efforts. Total
amount to be made available is not expected to exceed $2k.

ME Senior Project 2020 - 2021_Vantsevich, Zhang

You might also like