The following are last-minute news you should be aware of ;-)
* 05/08/2022: Grades of the 06/07/2022 call. * 21/07/2022: Grades of the Homeworks (updated) are out find them here!. * 20/07/2022: Grades of the Homeworks are out find them here!. * 07/07/2022: Final results of 15/06/2022 and 06/07/2022 for Laureandi. * 03/07/2022: Results of 15/06/2022 written exam are here. They include only the written grades, maximum is 26! * 03/06/2022: Updated all slides of the course, including last lecture * 02/06/2022: Last lecture of the course will be online Friday 03/06/2022 from 09:00 to 11:00 * 01/06/2022: Slides updates and new lecture rescheduling ongoing * 31/05/2022: Poll for the last lecture is here! * 19/05/2022: The second homework is out!!!! Don't forget to use the Slack channel for requests and suggestions!!! * 11/05/2022: Updated the detailed schedule of lectures, double check it!!! * 11/04/2022: Corrected bags and updated slides are now available in the shared folder. New deadline: May 8 2022, 23:59 CEST * 11/04/2022: All communications regarding the project will be though the shared folder and slack * 04/04/2022: Hold-on! -> due to a bug in the bags an update of the homework will be available soon * 30/03/2022: The first homework project of the course is out! * 06/03/2022: Added link to Paolo Cudrano slides, updated slide first two weeks Matteo Matteucci * 06/03/2022: Added link to last year's recordings * 23/02/2022: Detailed calendar published * 23/02/2022: Lectures start today!
- 1 Course Aim & Organization
- 2 Teaching Material (the textbook)
- 3 Frequently Asked Questions
- 4 Past Years Useful Material
Course Aim & Organization
This course will introduce basic concepts and techniques used within the field of autonomous mobile robotics. We analyze the fundamental challenges for autonomous intelligent systems when these move on wheels or legs and present the state of the art solutions currently employed in mobile robots and autonomous vehicles with a focus on autonomous navigation, perception, localization, and mapping.
The course is composed by a blending of lectures and exercises by the course teacher and a teaching assistant.
- Matteo Matteucci: the course teacher and this is his webex room
- Simone Mentasti: the teaching assistant and this is his webex room
- Paolo Cudrano: the teaching assistant and this is his webex room
Lectures will provide theoretical background and real-world examples. Lectures will be complemented with practical software exercises in simulation and on real data for all the proposed topics and the students will be guided in developing the algorithms to control an autonomous robot.
Among other topics, we will discuss:
- Mobile robots kinematics,
- Sensors and perception,
- Robot localization and map building,
- Simultaneous Localization and Mapping (SLAM),
- Path planning and collision avoidance.
Detailed course schedule
A detailed schedule of the course can be found here; topics are just indicative while days and teachers are correct up to some last minute change (I will notify you by email). Please note that not all days we have lectures!!
Note: Lecture timetable interpretation
- On Wednesday, in 26.11, starts at 12:15, ends at 14:15
- On Thursday, in 26.11, starts at 14:50, ends at 16:15
|23/02/2022||Wednesday||12:15 - 14:15||26.11||Matteo Matteucci||Lecture||Course/Robotics Intro|
|24/02/2022||Thursday||14:15 - 16:15||26.11||Matteo Matteucci||Lecture||Actuators|
|02/03/2022||Wednesday||12:15 - 14:15||26.11||Matteo Matteucci||Lecture||Sensors|
|03/03/2022||Thursday||14:15 - 16:15||26.11||Paolo Cudrano||Laboratory||ROS Install|
|09/03/2022||Wednesday||12:15 - 14:15||26.11||Matteo Matteucci||Lecture||Sensors|
|10/03/2022||Thursday||14:15 - 16:15||26.11||Paolo Cudrano||Laboratory||ROS Basics|
|16/03/2022||Wednesday||12:15 - 14:15||26.11||Matteo Matteucci||Lecture||Kinematics|
|17/03/2022||Thursday||14:15 - 16:15||26.11||Paolo Cudrano||Laboratory||Pub / Sub|
|23/03/2022||Wednesday||12:15 - 14:15||26.11||Matteo Matteucci||Lecture||Kinematics|
|24/03/2022||Thursday||14:15 - 16:15||26.11||Paolo Cudrano||Laboratory||Services and Params|
|30/03/2022||Wednesday||12:15 - 14:15||26.11||Matteo Matteucci||Lecture||Localization and Lidars|
|31/03/2022||Thursday||14:15 - 16:15||26.11||Paolo Cudrano||Laboratory||TF / Rviz / Actions|
|06/04/2022||Wednesday||12:15 - 14:15||26.11||Matteo Matteucci||Lecture||Localization and Bayes Filters|
|07/04/2022||Thursday||14:15 - 16:15||26.11||Matteo Matteucci||Lecture||Localization and Kalman Filters|
|13/04/2022||Wednesday||---||---||---||---||--- Prove in itinere ---|
|14/04/2022||Thursday||14:15 - 16:15||26.11||Matteo Matteucci||Lecture||Localization and Kalman Filters|
|20/04/2022||Wednesday||12:15 - 14:15||26.11||Matteo Matteucci||Lecture||Localization and Kalman Filters|
|21/04/2022||Thursday||14:15 - 16:15||26.11||Simone Mentasti||Laboratory||rospy, rosbag, message filter, plotjuggler|
|27/04/2022||Wednesday||12:15 - 14:15||26.11||Matteo Matteucci||Lecture||Localization and Kalman Filters|
|28/04/2022||Thursday||---||---||---||---||--- Lauree ---|
|04/05/2022||Wednesday||12:15 - 14:15||26.11||Matteo Matteucci||Lecture||Mapping and SLAM|
|05/05/2022||Thursday||14:15 - 16:15||26.11||Simone Mentasti||Laboratory||ROS on multiple machines, time synchronization, Actionlib, latched pub, async spinner|
|11/05/2022||Wednesday||12:15 - 14:15||26.11||Matteo Matteucci||Lecture||Mapping and SLAM|
|12/05/2022||Thursday||14:15 - 16:15||26.11||Matteo Matteucci||Lecture||Algorithms for Robot Navigation|
|18/05/2022||Wednesday||12:15 - 14:15||26.11||Simone Mentasti||Laboratory||Robot Navigation, Stage, Gmapping|
|19/05/2022||Thursday||14:15 - 16:15||26.11||Simone Mentasti||Laboratory||Robot Navigation (Part II), Robot Localization, mapviz|
|25/05/2022||Wednesday||---||---||---||---||--- No Lecture ---|
|26/05/2022||Thursday||14:15 - 16:15||26.11||Simone Mentasti||Laboratory||ROS2, foxglove, second project|
|01/06/2022||Wednesday||12:15 - 14:15||26.11||Matteo Matteucci||Lecture||Algorithms for Robot Navigation|
|02/06/2022||Thursday||---||---||---||---||--- No Lecture ---|
|03/06/2022||Friday||09:00 - 11:00||Online||Matteo Matteucci||Lecture||Algorithms for Robot Navigation|
Course evaluation is composed by two parts:
- A written examination covering the whole program graded up to 26/32
- A home project in simulation practicing the topics of the course graded up to 6/32
The final score will sum the grade of the written exam and the grade of the home project.
Course Project (i.e., the two  homeworks)
In the course project, you will use ROS to develop a simple autonomous mobile robot performing simple mapping, localization, and navigation task. The project requires some coding either in C++ / Python following what will be presented during the lectures (we suggest using C++ as it will be the language used in class). The project will be presented in two (2) parts you have about one month to do each. Details will follow.
Teaching Material (the textbook)
Lectures will be based on material from different sources, teachers will provide their slides to students as soon they are available.
Course Slides 2021/2022
Slides from the lectures by Matteo Matteucci
- [2021/2022] Course Introduction: introductory slides of the course with useful information about the course syllabus, grading, and the course logistics.
- [2021/2022] Introduction to Robotics: Introduction to Robotics, definitions, examples and SAP cognitive model.
- [2021/2022] Sensors and Actuators: an overview of most commonly used actuator and sensors in robotics, the DC motor and its characteristics, gears and torque. Sensor classification, common sensors in robotics with pros and cons.
- [2021/2022] Robot Odometry: Robot Localization intro, direct and inverse kinematics, robot odometry for different kinematics (differential drive, skid steering, Ackerman, etc.).
- [2021/2022] Robot Localization: Sensor models, Robot Localization, Bayesian filtering, Kalman Filtering, Monte Carlo Localization.
- [2021/2022] Simultaneous Localization and Mapping: Mapping with known poses, scan matching, EKF-SLAM, FAST-SLAM
- At this link you can find the videos included in the slides about (simulataneous) localization and mapping
- [2021/2022] Robot Motion Control: Introduction to motion control, Virtual Histogram methods, Dynamic Window Approach, Search-based and Sampling-based planners (here the video recording of last lecture)
Slide from the teaching assistant, including code examples and homework descriptions are available in their respective folders
- Last version of slides from the lectures by Paolo Cudrano are available HERE!.
- Last version of slides from the lectures by Simone Mentasti are available HERE!.
Year 2020/2021 Recording
As I registered these due to pandemics, I am making them available. They DO NOT REPLACE THIS YEAR classroom lectures which are to be considered as the official material of this year, but they might be useful to double-check your notes.
- 24/02/2021 - Matteo Matteucci Course logistics + Introduction to Robotics
- 03/03/2021 - Matteo Matteucci Sensors and Actuators
- 10/03/2021 - Matteo Matteucci Robot Sensors and Intro to SLAM
- 17/03/2021 - Matteo Matteucci Robot Kinematics (Differential Drive)
- 31/03/2021 - Matteo Matteucci Robot Kinematics (Continued)
- 07/04/2021 - Matteo Matteucci Localization and LiDARS
- 14/04/2021 - Matteo Matteucci Localization and Bayes filters
- 21/04/2021 - Matteo Matteucci Localization with Kalman filters and Particle filters
- 05/05/2021 - Matteo Matteucci Mapping and SLAM
- 12/05/2021 - Matteo Matteucci Robot Motion Control
- 26/05/2021 - Matteo Matteucci Search-based Planning
- 31/05/2021 - Matteo Matteucci Sampling-based Planning
Also, labs are available, however, the organization during the pandemic was kind of different with 2 teams to reduce classroom occupancy. This is why they might resemble a kind of disconnected.
- 02/03/2021 - Paolo Cudrano Introduction to middleware in Robotics
- 09/03/2021 - Paolo Cudrano ROS Basics
- 17/03/2021 - Paolo Cudrano Publishers and Subscribers
- 24/03/2021 - Paolo Cudrano Services and Parameters
- 31/03/2021 - Paolo Cudrano TF, RVIZ (Cudrano recording)
- 13/04/2021 - Paolo Cudrano Bags, Message filters and rospy + Project 1 Presentation
- 04/05/2021 - Simone Mentasti ROS on Multiple Devices
- 12/05/2021 - Simone Mentasti Robot Navigation
- 26/05/2021 - Simone Mentasti IMU Tools and robot localization + Project Presentation + Project folder
Useful stuff from the web
These are videos from the web which might be useful to understand better the material presented in the lectures
- Understanding DC Brushed Motors by Learn Engineering
- Understanding DC Brushless Motors by Learn Engineering
- Understanding DC Stepper Motors by Learn Engineering
This blog post can be useful to better understand the EKF-SLAM idea and implementation
- EKF-SLAM hands-on tutorial by Jihong Ju
If you have problems in installing Linux on your machine you can use a USB drive distro and boot on it instead of your OS. Note: We are testing this guide these days we might have some tips and tricks for it so stay tuned!
- How to install a complete ubuntu on a USB flash drive (need to have the USB drive inserted to boot)
The ROS framework is C++ based, if you want to check some C++ tutorial online you can have a look at
- Simple, basic topics about C++
- A more detailed tutorial about C++
- An even more detailed tutorial on C++ (you can just focus on some particular chapters. In particular, Ch. 11 seems interesting as a detailed overview of Object-Oriented Programming, if you are not familiar with it.)
These are papers which explain some of the topics in the lecture with a higher level of details
- Analysis and experimental kinematics of a skid-steering wheeled robot based on a laser scanner sensor. Wang, Tianmiao, Yao Wu, Jianhong Liang, Chenhao Han, Jiao Chen, and Qiteng Zhao. Sensors 15, no. 5 (2015): 9681-9702.
- Simultaneous localization and mapping with the extended Kalman filter. Joan Sola'.
- FastSLAM: An Efficient Solution to the Simultaneous Localization And Mapping Problem with Unknown Data Association. Sebastian Thrun, Michael Montemerlo, Daphne Koller, Ben Wegbreit, Juan Nieto, and Eduardo Nebot.
- The Dynamic Window Approach to Collision Avoidance. Dieter Fox, Wolfram Burgard, and Sebastian Thrun.
- The First project con the Robotics class is available HERE!, deadline is 29/04/2022!
- The Second project con the Robotics class is available HERE!, deadline is 26/06/2022!
Frequently Asked Questions
What is the biggest difference with the course 093217 ROBOTICS AND DESIGN?
- Robotics and Design is a practical course focused on the development of a robotics application, you will not learn about the theoretical aspects of robotics, but you will build a robot with a purpose which changes every year. I consider the two courses as complimentary.
Exams and Evaluation
Are there any solutions available for the past exams?
- No, if you have doubts or questions, just send me your proposed solution and I will reply tailoring the answer to your current understanding.
Is it important to buy/read the text book to be able to follow the course? I can’t find it in the library, is there any alternative book?
- No, it is not required, as from past experience attending classes and checking the material provided y the teachers is enough. Obviously reading the book will provide much more information..
Homeworks and ROS
In the schedule when it says ROS, are these lectures as well or are they practical work i.e. lab/excercise?
- They are ex-cathedra lectures where you are expected to bring your laptop, it is not mandatory and you can follow the class in a classical passive way, but I suggest to consider it as a lab and take your laptop with you if you can.
Out of all the scheduled activities this semester, approximately how many of these are practical lab/excercise?
- Indeed not all ROS lectures will present coding exercises, I expect half of them will be about coding and the other half more on the technical background you need to understand what you are coding.
Should I install ROS on my laptop/desktop?
- Absolutely yes. This means you need to have linux on your machine, possibly ubuntu 16.04 or 18.04. This can be achieved in different ways, we suggest a native install via dual boot or as main operating system (we do not take any responsibility of something happening to your data or hardware in doing this operation). Other options such as virtual machine or live distro are not as effective as a real install, but they work.
Which editor/IDE should I use for ROS?
- We do not suggest any particular editor for ROS, standard text editors such as nano/gedit/sublime + a terminal are enough. Nevertheless, you can use the environment you prefer for C++ development; some students, in the past, have used Eclipse or Clion. You can also check the list of supported ROS editors or Roboware, the latter has been designed for ROS, but it does not offer any special feature you will miss using standard C/C++ editors.
As I understand the “homework/project” is a group project. Is this correct and how are the groups formed?
- It is not a group project, while it is allowed to do it in groups (up to 3 people). I expect the groups to form naturally in classes. We usually set up a slack group for the project you can organize autonomously. Nevertheless, you can do the project alone as well (but we advise you to do it in groups).
When “Part 1” of the homework/project will start?
- Right after we have finished the first block of lectures about ROS. This should happen around Easter plus/minus a week.
Past Years Useful Material
Here you find material from past editions of the course that you might find useful in preparing the exam.
Past Exams and Sample Questions
Since the 2015/2016 Academic Year the course has changed the teacher and this has changed significantly the program and the exam format as well. For this reason we do not have many past exams to share with you, they will accumulate along the years tho.
- Exam of 17/07/2017
- Exam of 01/07/2017
- Exam of 26/09/2016
- Exam of 05/09/2016
- Exam of 20/07/2016
- Exam of 27/06/2016
Note on 06/07/2022 Exam
This is a short note on the grading of the 06/07/2022 exam. On average has not been different from the others calls except one exercise you might want to know more about, i.e., exercise 1. I take this opportunity to comment on the grading of all exercises so you can get an immediate comment and if there is something missing you can then write to me.
- Exercise 1: The key point here is the use of a "single RGB camera", because of this, we do not have distance measurements and cannot consider the sensor as lidar, sonar, or stereo vision system. Because of this, beam or scan sensor models dedicated to range sensors are not applicable. In this case, you need to use a landmark-based sensor model getting landmarks from the vision system, e.g., a door or a fire extinguisher (examples made multiple times during lectures). As for the localization system any solution which leverages this sensor model is ok, it could be an Extended Kalman Filter, or a Monte Carlo Localization (as the soccer dog example we made in class and you find in the slides), but what is important is to motivate the system on the characteristics of the landmark sensor, not just provide a generic description of a localization algorithm.
- Exercise 2: More or less all exams got the first two points correct, as for the third two options equally correct exist. A) you specify the actuator on the back wheels for forward motion and on the frontal wheels for steering and then you provide the derivation of the Ackerman kinematics; B) you specify that frontal wheels are just caster wheels for support and then you have 2 independent motors on the back deriving then the differential drive kinematics. Partial answers get partial grades. If you are curious the "real" robot is made like B)
- Exercise 3: the decision on the action to choose is based on a scoring function which is different in the cases of 3.1, 3.2, and 3.3. Describing these scoring functions provides full marks, only stating they exist or mixing them up only partial mark, not mentioning these scoring functions zero mark. As for 3.4, the solution is manually (or automatically) tuning the coefficients via trial and errors possibly in simulation.
- Exercise 4: not stating clearly the characteristics of topics, services and actions gives partial credit.
- Exercise 5: most of you got the point, full SLAM requires to estimate the map and the full trajectory. If you want to implement it using an EKF-SLAM algorithm you need to add all the poses to the state of the EKF and estimate jointly the map and the trajectory. This makes the state grow linearly with time and the complexity quadratically with the size of the state ... if you are curios you can check SAM (Smoothing and Mapping) id does exactly this.
- Exercise 6: who has applied A* got full mark, who has just searched for the path and found it by looking at the graph exploring it partially got very low mark. Intermediate marks are give because of incomplete solutions or errors in the execution.
Past Course Project
Here you find past course projects in case you are interested in checking what your colleagues have been pass through before you. In some cases they may have been more lucky in some others you might be the lucky one ... that's life! ;-)
Here they are the curse homework projects:
- The first course project has been published on 14/04/2021
- The second course project has been published on 26/05/2021
Here they are the curse homework projects:
The 2018/2019 course project is divided in two releases. The homework philosophy should be "You have to struggle, but not too much!". Indeed the homework is made to challenge you and make you exercising and learn by doing, nevertheless if you find yourself stuck please write us and we will give you the required hints to continue and complete ... this includes extending the deadline (for all) or allowing you to use python instead of C++ (for selected students).
Advice: Start as soon as possible doing the homework!
- 2018/2019 Course Project Part 1: due on Wednesday 29/05/2019, this is the first part of the 2018/2019 course project.
- 2018/2019 Course Project Part 2: due on Monday 08/07/2019, this is the second and last part of the 2018/2019 course project.
The 2016/2017 course project is divided in two releases to provide you something to work on as early as possible during the course. The homework philosophy should be "You have to struggle, but not too much!". Indeed the homework is made to challenge you and make you exercising and learn by doing, nevertheless if you find yourself stuck please write us and we will give you the required hints to continue and complete.
Advice: Start as soon as possible doing the homework!
- 2016/2017 Course Project Part A v1.1: due on Wednesday 31/05/2017 (6 weeks from now), this is the first part of the 2016/2017 course project.
- 2016/2017 Course Project Part B v1.0: due on Wednesday 28/07/2017 (6 weeks from now), this is the second part of the 2016/2017 course project.
- 2016/2017 Model for Course Project part B v1.0: thi si the gazebo model to be used in exercise 4 in the second part of 2016/2017 course project.
This year project is divided in steps; each of them is worth some points out of the 5/32 points available for the final mark. You find the project description here, it is complete, it contains parts up to 4, parts 5 is optional, but we suggest to do it anyway since it requires a limited amount of time.:
- 2015/2016 Course Project v1.0
- 2015/2016 Kobra STL files: in case you want to make your simulation look more real here you find the STL files of the Kobra robot in the "Safer" version. Unfortunately the STL files are scaled down with respect to the real robot, so you have to modify those if you want to use.
If you are interested in a more deep treatment of the topics presented by the teachers you can refer to the following books and papers:
- Probabilistic Robotics by Dieter Fox, Sebastian Thrun, and Wolfram Burgard.
The following are links to online sources which might be useful to complement the material above
- ISO 8373:2012: ISO Standard "Robots and robotic devices -- Vocabulary"
- ROS: the Robot Operating System
- Gazebo: the Gazebo robot simulator
- AIRLab ROS Howto: a gentle introduction to ROS with node template and program examples