ELEC330
Assignment Two
Robot Moving in its World and Face Detection
Dr H LAKANY
October 30, 2023
Abstract
This assignment has a weight of 30% of the total grade. The assignment
builds on your first assignment and uses the robot you have designed. The
assignment is issued on the fifth week and is due on Friday of week twelve
of the first semester. This assignment is very important as it represents a
stepping stone to enable you to complete the third and final assignment.
Warning: When marking the reports I will be looking very closely for
any signs of collusion, as this compromises your academic integrity. This is
a serious offence. I need to assess your own ability not that of your friend or
colleague. If I found any evidence of collusion then the University’s rules on
the punishment for collusion will be followed. If you do collude I will spot
it and you will be penalised – you have been warned.
1
1 Assignment requirements and marking criteria.
Table 1 summarises the assignment requirements.
Table 1: Assignment Requirements
Module ELEC330
Coursework title Assignment 2
Component weight 30%
Semester 1
HE level 6
Lab location PC labs 301,304 as timetabled and private study
Work Individually
Timetabled time 24 hours (3 hours per week, Wednesdays 1000-1300 )
Suggested private study 16 hours including report writing
Assessment method Individual, formal word–processed reports in addition
to video and code.
Submission format On–line via CANVAS
Submission deadline 23:59 on Friday 15th December 2023
Late submission Standard university penalty applies.
Resit opportunity Students failing Assignment 1 will have Assignment 2
as a resit opportunity.
University Rules will be applied if total module failed
Marking policy Marked and moderated independently
Anonymous marking No
Feedback individual feedback issued.
Learning Outcomes LO1–LO3: Describe the main components of a robotic system.
Design, model and simulate a simple robotic system.
LO5: Describe and implement sensing in a robotic system.
LO6: Define and implement a system for navigation of a
robotic system.
LO7: Explain how machine vision, machine learning and
human machine interfaces augment capability of robotic
systems.
2
Table 2 summarises the marking criteria of the second assignment.
Table 2: Marking Criteria
Section
Marks
available
Indicative Characteristics
Adequate/pass (40%) Very good/Excellent
Clarity of
presentation
(including
succinctness) of
main report
20%
2–page executive summary.
Comprehensible language
with adequate grammar and
punctuation. Documented
code and Gazebo .world files
in appendices
Appropriate use of technical
language. All sections clearly
signposted. Correct cross
referencing to appendices.
Annotated Gazebo .world
and code files including
readme files.
Methodology and
testing scenarios
40%
Adequate Gazebo .world file;
Rviz, minimum testing
scenarios of the robot
designed in assignment 1
navigating autonomously to
target.
Well thought through .world
file and testing scenarios
foreseeing requirements of
the exploration mission and
detecting face of at least one
human avatar in the indoor
environment. Video showing
the detail of the testing of
various navigating scenarios
and face detection schema.
Quality of
discussion,
reflection and
conclusion
40%
Critical assessment of the
efficacy of the robot in
navigating the world and
identifying the target.
Discussion on what worked
and what did not.
Discussion how the robot
successfully navigated its
world, detected the face of
the human avatar and how
the model was fully
assessed/tested.
3
2 Assignment Outline
In this assignment, you will
• create a .world file in which the robot you designed in the first assign-
ment will navigate.
• add at least two sensors suitable for creating a map of the world.
• use gmapping to create a .map file of the world.
• test different navigation scenarios.
• create/include a human avatar and spawn in the indoor environment.
• be using the navigation stack to implement autonomous navigation.
• drive the robot autonomously to a human model.
• detect the face of the human and take a photo using a camera.
In this assignment, you shall start by creating and launching a world file
in Gazebo. The world file will represent an indoor space that has at least
four walls, one door and one window and has the following objects on its
floor:
1. at least one chair,
2. at least one coffee table,
3. one book shelf,
4. at least one sphere,
5. at least one cube,
You can use objects from the Gazebo database.
There are no requirements on
• the dimensions of the indoor space but should be large enough to enable
the navigation of your robot within it.
• the position of the various objects and the distances among them.
• the starting position of the robot.
After creating the .world file, you will choose an appropriate sensor to
add to your robot to enable it to generate a map of the world using gmap-
ping (SLAM) and use YAML to load it.
You should be able to use both Rviz for visualisation and Gazebo for
simulation.
4
You may refer to the following tutorials that you were previously asked
to study including:
• http://wiki.ros.org/rviz to refresh your knowledge in Rviz (ROS
visualisation tool)
• http://gazebosim.org/tutorials?cat=build_world. to build a world.
• http://wiki.ros.org/Robots/TIAGo/Tutorials/Navigation/Mapping.
• http://wiki.ros.org/map_server.
make sure you are using the correct distribution e.g. Melodic or Noetic
version.
• https://emanual.robotis.com/docs/en/platform/turtlebot3/slam/
#ros-1-slam.
• http://wiki.ros.org/amcl to localise the robot within the indoor
space.
• http://gazebosim.org/tutorials?tut=ros_gzplugins.
You should be able to simulate the model in Gazebo and visualise it in Rviz.
You need to SLAM/map the Gazebo world to make its robot perception
equivalent available in Rviz. So the robot needs to ”scan” the Gazebo
world, translate it into map format, and publish it (via map server)
5
3 Submission
3.1 Submission Deadline
The submission deadline for the assignment is Friday 15th December 2023
at 23:59. You will submit your report and code on–line using the dedicated
link in the module page on CANVAS.
3.2 What to submit
You will submit an executive summary, a detailed report, code and a video
as follows. Use appropriate names for all files and zip them in one file to
submit.
3.2.1 Executive Summary
You will submit a 2–page executive summary with a maximum of two fig-
ures (excluding any appendices) in 10 point font with no less 2cm margins.
Note that this summary is very important and will be the only part that
is completely read by the marker. It should include all what is needed to
get the reader acquainted of what you have achieved and how. The reader
should not have to refer to any other material. The executive summary
should include
• Summary of the world you designed with a bird’s eye view of the world.
• Selection of the sensor and any modifications to the initial robot design.
• Autonomous navigation and face detection.
• Discussion and reflection on (1) map generation, (2) testing scenarios
to generate an accurate map, (3) accuracy of the map generated with
respect to the robot starting position.(4) localisation and navigation,
(5) testing scenarios to generate an accurate map and reach the target
position, (6) image of the human face.
• Conclusion with a view on what requirements are needed for a mobile
robot to accurately detect objects including human faces.
3.2.2 Appendices
You should use appendices where you include the detail of all the steps you
have gone through to complete the creation of the world and generate the
map. The appendices should include a detailed section for each of require-
ment specification of the world, launch file, the map, the testing scenarios,
how you achieved the an accurate map. A reflection section on what worked,
what did not work, what are the lessons learnt and skills acquired.
6
3.2.3 code
This will include any code you have written or used to complete the assign-
ment. This will also include all the .world, .yaml and any other files. All
code and files should be appropriately documented and annotated.
3.2.4 video
Submit a video showing the robot in Gazebo and Rviz navigating and iden-
tifying the human avatar. Figures will also be accepted.
4 How to tackle the assignment?
To help you tackle the assignment, a series of laboratory sheets are pub-
lished weekly to help you work on your assignment step-by-step. The first
laboratory sheet associated with the second assignment is Lab Sheet 04. It
is your responsibility to complete the tasks in the laboratory sheets which
will help you to successfully complete the assignment.
5 Difficulties
If you have difficulty
• designing and implementing an indoor environment in Gazebo, you
can use an existing .world file and modify it. If you are unable to do
that, you can use .world file without modification.
• using Makehuman to create human models, you can use an existing
person model in Gazebo object database.
• in autonomously driving the robot, then you can use keyboard control
to drive it to target.
• to detect a face using a camera, take a photo of another object in the
room.
Taking any of these easy solutions will affect your grade.