Robotics 1
Introduction
Prof. Alessandro De Luca
Robotics 1 1
Robotics 1 – 2024-25
n First semester
n Friday, September 27, 2024 – Friday, December 20, 2024
n Courses of study (those having Robotics 1 mandatory or as optional)
n Master in Artificial Intelligence and Robotics (MARR)
n Master in Control Engineering (MCER)
n 6 Credits
n ~50 hours of lectures, exercises, and midterm test
n 90 hours of individual study
n Classes (room B2, DIAG, Via Ariosto 25)
n Wednesday 17:00-19:00
n Friday 8:00-11:00
Robotics 1 2
General information
n Prerequisites
n self-contained course, without special prerequisites
n elementary knowledge on kinematics (of a point, of a single
rigid body), linear algebra (vectors and matrices), and
feedback control (of linear dynamical systems) is useful
n Aims
n robot “anatomy”
n provide tools for kinematic analysis of articulated chains of
multiple rigid bodies (= robot manipulators)
n analytical methods for planning motion trajectories
n motion command and control algorithms
n programming of tasks for industrial & service manipulators
Robotics 1 3
Organization and contacts al l
are hyperl
acti i
ve i nks in
n th r
e pd ed
f f i le
n G-group
n join robotics1_2024-25 (your full name, uniroma1 email, course of study)
n ask questions, comment on my replies (for the benefit of everyone!)
n Email [email protected]
n Office hours for students
n Tuesday 12:00-13:30 (check exceptions, e.g., when I’m on travel)
n in presence: Room A-210, floor 2, left wing, DIAG, Via Ariosto 25
n remote: Zoom (or Meet), see www.diag.uniroma1.it/deluca/Teaching.php
Robotics 1 4
A robot manipulator
Illustrating typical features of an industrial robot
commercial video
KUKA KR 4 Agilus robot with 6 revolute joints
Robotics 1 5
Robot manipulators
available at DIAG Robotics Lab (S-218)
video video
KUKA KR-5 KUKA LWR4+ (lightweight, about 14 kg)
Robotics 1 6
Robot manipulators
available at DIAG Robotics Lab (S-218)
commercial video
upon arrival (July 2016) Universal Robots UR-10 (= 10 kg of payload)
DIAG Robotics Lab YouTube channel for more of our research videos
www.youtube.com/user/RoboticsLabSapienza
Robotics 1 7
Programming robot motion
Teaching Cartesian poses and playing them back
video
KUKA LBR iiwa robot with 7 revolute joints
Robotics 1 8
Programming robot motion
Executing nominal trajectories and “complying” with uncertainties
video
actual
compliance
nominal
Robotics 1 9
Programming robot compliance
Controlled reaction to applied forces/torques at robot end-effector
video
2000 N/m
300 N/m
5000 ⇒ 300 ⇒ 100 N/m ∆𝑝 → 𝐹 = 𝐾 ∆𝑝
stiffness
vs
compliance
10 N/m
𝐹 → ∆𝑝 = 𝐶 𝐹
300 Nm/rad
1
= 𝐹
𝐾
5000 N/m
10 Nm/rad
Robotics 1 video 10
Programming robot motion
Teaching tasks by demonstration (kinesthetic learning)
video
sketch of the original idea
— a possible use of safe
physical Human-Robot Interaction (pHRI)
video
the working industrial solution
more videos on
KUKA Robotics YouTube Channel
Robotics 1 11
Program
n Introduction
n Manipulator arms (+ some on a mobile base)
n Industrial and service applications
n Components
n Mechanical structures
n Actuators and transmissions
n Sensors
n proprioceptive (encoders, velocity reconstruction, …)
n exteroceptive (force/torque, depth, vision, infrared, ultrasound, laser)
n Kinematic models
n Minimal representations of orientation
n Direct and inverse kinematics of robot manipulators
n Denavit-Hartenberg formalism for frame assignment
n Differential kinematics: analytic and geometric Jacobians
n Statics: Transformations of forces
n Robot singularities
Robotics 1 12
Industrial vs. service robots
FANUC CR35i 6R collaborative robot
carrying up to 35 kg of payload
and with hand guidance system
video
video
qbrobotics SoftHand2
dexterous robot hand
with only 2 motors on board
and compliant transmissions
Robotics 1 13
Program (continued)
n Planning of motion trajectories
n Trajectory planning in the joint space for robot manipulators
n Trajectory planning in the task/Cartesian space
n Motion control
n Control system architectures
n Kinematic control laws (in joint or in task/Cartesian space)
n Independent joint axis control laws (P, PD, PID)
n Programming and simulation
n Programming language for industrial robots (KRL)
n Use of Matlab/Simulink (with Robotics Toolbox) or CoppeliaSim (V-REP)
Robotics 1 14
Planning Cartesian trajectories
video
KUKA 6R articulated robot 2R planar robot
t = [0:Ts:T]; tau = t/T; s = (-2*tau.^3 + 3*tau.^2); center = [0.5; 0.5];
p = center + [(16*sin(2*pi*s).^3)/16;
(13*cos(2*pi*s) - 5*cos(4*pi*s) - 2*cos(6*pi*s) - cos(8*pi*s))/16]/2;
Robotics 1 MATLAB code (Diaco & Ficorilli, Sep 2021) 15
Programming motion and behavior
video
myBuddy 280-Pi dual-arm personal robot by Elephant Robotics
Robotics 1 16
Tracking a Cartesian trajectory
with hard position/velocity bounds on robot motion
video DIAG Sapienza/Stanford, IEEE ICRA 2012
Robotics 1 17
Robot control by visual servoing
with limited joint motion range
video INRIA Rennes, IEEE/RSJ IROS 2011
Robotics 1 18
Sensor-based robot control
in dynamic environments (coexistence with human)
video DIAG Sapienza/Stanford, IEEE ICRA 2012
Robotics 1 19
Safe physical human-robot interaction
(sensor-less (!) and on a conventional industrial robot)
video DIAG Sapienza, IEEE ICRA 2013
Robotics 1 20
Human-robot collaboration
(with a real F/T and a “virtual” sensor to distinguish contacts)
video DIAG Sapienza, J. of Mechatronics, 2018
Robotics 1 21
More robots …
@IEEE https://robotsguide.com
KUKA LBR iiwa
7-dof arm
DE 2013
UNIVERSAL ROBOTS
UR5 6-dof arm
DK 2008
https://youtu.be/7GdiN6KmGCc https://youtu.be/UQj-1yZFEZI
ABB YuMi-FRIDA
bi-manual arm
SE-CH 2011
OCRobotics Explorer
snake arm
UK 2009
https://youtu.be/70V6J4Y8hnc https://youtu.be/_gU6TWGynkU
Robotics 1 22
Next generation of intelligent robots?
Robots = cognitive robotics
embodied AI systems cyber-physical systems
mechanics,
electronics,
Robotics =
AI Robotics hydraulics,
science
…
of artifacts
intelligently
learning,
actuated and
adaptation,
interacting with
system
the physical world
identification
(data-driven)
model-based
techniques, robot motion
dynamic primitives, and interaction
uncertainty Control control
Engineering
Robotics 1 23
AI and Robotics
manipulation tasks
Toyota Research Institute
(work by Russ Tedrake)
dishwasher loading:
a manipulation task (speeded up!)
that integrates planning, vision,
learning and control
video
video
pick-and-place motion:
programming a manipulation task
from LLM commands
(using a library of motion primitives!)
Microsoft
ChatGPT for Robotics
Robotics 1 24
Levels of robot autonomy
in industrial applications
THE FIVE LEVELS OF AUTONOMY IN ROBOTS
Pick-and-place robot
Collaborative Assistant robot that
Remote-operated Welding robot in feeding a machine
assembly robot in responds differently to
surgical robot automotive factory with parts from an
manufacturing a child than an adult
unsorted bin
source, IFR 2022
Robotics 1 25
Horizon Europe (2021-27)
Cross-Sectorial AI, Data and Robotics Technology Enablers
Sensing and Perception
Knowledge and Learning
Reasoning and Decision Making
Action and Interaction
Systems, Methodologies, Hardware & Tools
Robotics Deep Dive
Physical Interaction
Physical and Psychological Safety
Actuated Mechanical Structures
Unpredictable and Unknown Environments
Irreversible Actions
4 Market Prioritization in Robotics (from Horizon 2020)
Healthcare
Maintenance and Inspection of Infrastructures
Agri-Food
local link to the document
Agile Production
Robotics 1 26
Specific to Robotics
§ hard physical/real-world nature of Robotics (“AI embodied”)
§ wide range of technologies are integrated within robotic systems
§ skill mix needed for success is broader than in AI or Data alone
§ robots are realizations of advanced system-level concepts
§ such as autonomy, control, sensing, perception and programming
§ robots are both producers and consumers of data
§ physical model-based approaches, generation of data-driven models
§ decision makers and general public need a better understanding of
what Robotics is and can achieve, and how it can be deployed
§ Fukushima, Covid19, ….
Robotics 1 27
AI & Robotics point of views
example with one of the cross-sectorial technology enablers
one … its relations with the other technologies
enabling
technology
Safe control of physical
human-interaction
Real-time interpretation Active exploration strategies Planning and re-planning Agility (speed and strength)
of multi-modal data under uncertainty and of collaborative robots
Adaptive decision-making
Safe monitoring in models from sparse data incomplete knowledge InterAction Technology
human environments in dynamic environments (lightweight, compliant
Real-time control & soft devices/materials)
(distributed/decentralized) Energy-efficient, robust
Robotics 1 and sustainable design 28
Course materials
Textbook
Bruno Siciliano Bruno Siciliano
AUTOMATICA
Lorenzo Sciavicco
L u i g i Vi l l a n i Lorenzo Sciavicco
n
Giuseppe Oriolo L u i g i Vi l l a n i
Giuseppe Oriolo
Robotica
L . Vi l l a n i • G . O r i o l o
B. Siciliano • L. Sciavicco
Modellistica, pianificazione e controllo
B. Siciliano, L. Sciavicco, L. Villani, G. Oriolo:
Questo libro si pone in continuità con le due precedenti
edizioni del testo di Robotica Industriale, a firma dei primi
due autori, di cui mantiene l'obiettivo didattico di coniu-
gare gli aspetti fondamentali e tecnologici con quelli inno-
n
vativi attraverso una presentazione formalmente rigorosa.
I primi sei capitoli del testo riguardano la teoria delle
strutture di manipolazione (con la trattazione della cine-
matica, della statica e della pianificazione di traiettorie) e la
Robotics: Modelling, Planning and Control,
tecnologia degli attuatori, dei sensori e dell'unità di gover-
no di un robot; gli altri sei capitoli approfondiscono la
pianificazione e controllo
Modellistica,
Robotica
dinamica e il controllo del moto dei manipolatori, l'intera-
zione con l'ambiente con l'uso di informazioni sensoriali
esterocettive (forza e visione), i robot mobili e la pianifica-
Bruno Siciliano è professore zione del moto. Completano il volume cinque appendici
Robotica
di Automatica presso l’Università che forniscono una base di conoscenza omogenea a studen-
di Napoli Federico II.
Lorenzo Sciavicco è professore
ti di diversa estrazione.
di Automatica presso l’Università L'organizzazione degli argomenti, presentati con un
3rd Edition, Springer, 2009
di Roma Tre. crescente livello di difficoltà, consente l'adozione del libro
Luigi Villani è professore di come testo di riferimento sia per un corso al terzo anno del-
Automatica presso l'Università di
Napoli Federico II.
le lauree di primo livello in ingegneria automatica, elettro- Modellistica, pianificazione e controllo
Giuseppe Oriolo è professore
nica, gestionale, informatica e meccanica, sia per un corso
delle corrispondenti lauree di secondo livello, o anche per
edizione
Te r z a
di Automatica presso l’Università
di Roma La Sapienza. corsi monografici a livello dottorale.
Te r z a e d i z i o n e
McGraw-Hill
English, Italian, Chinese & Greek editions
6322-2
ISBN 978-88-386-6322-2
,!7II8D8-ggdccc!
n
! www.mcgraw-hill.it
€ 37,00 (i.i.) ! www.ateneonline.it
n Course website www.diag.uniroma1.it/deluca/rob1_en.php
n pdf of lecture slides ready (with some updates during the course)
n all videos shown during lectures (in zipped folders by block of slides)
n written exams (most with solutions), syllabuses, extra documents, …
n Video DIAG channel playlist Robotics 1 with full course of 2014-15
n 30 (+1 index) videos in classroom (≅ 41h, >141K independent views)
n YouTube personal channel: playlists with videos of selected lectures
recorded during the pandemic (2019-20 and 2020-21)
Robotics 1 29
… a new textbook is coming!
B. Siciliano, L. Villani, G. Oriolo, A. De Luca: Foundations of Robotics,
Springer, to be published in the spring of 2025
Bruno Siciliano
Luigi Villani
Giuseppe Oriolo
Alessandro De Luca
Foundations of Robotics
Robotics 1 30
Exams and beyond
n Type midterm test (about mid of November) + written exam
n Schedule of 2024-25 sessions
n 2 sessions at the end of this semester
n between January 7 and February 25, 2025 will open
on infostud
n 2 sessions at the end of next semester
in early December
n between June 3 and July 25, 2025
check the
n 1 session after the summer break course website!
n between September 1 and 23, 2025
n 2 extra sessions only for students of previous years, part-time, ...
n March 17−April 18 and October 8−November 6, 2025
n Signing up to exams
n on infostud (up to one week before the date of the written exam)
n Master theses
n samples at DIAG Robotics Lab www.diag.uniroma1.it/labrob/theses
Robotics 1 31
Preview of Robotics 2
6 credits, II semester, year 1
n Advanced kinematics / Robot dynamics
n Calibration
n Redundant robots
n Dynamic modeling: Lagrange and (recursive) Newton-Euler methods
n Identification of dynamic parameters/coefficients
n Geometrically constrained dynamics
n Control techniques
n Free motion linear/nonlinear feedback control, robust control, adaptive
control, iterative and online learning
n Constrained motion admittance, impedance, hybrid force-motion control
n Visual servoing (kinematic approach)
n Special topic
n Diagnosis and isolation of robot actuator faults
Robotics 1 32
Other courses on Robotics and Control
n Autonomous and Mobile Robotics (6 credits), I semester, year 2
n kinematics, planning, control of wheeled and legged mobile robots
n motion planning with obstacles, navigation, and exploration
n Prof. Oriolo www.diag.uniroma1.it/oriolo/amr
n Medical Robotics (6 credits), II semester
n robot surgical systems, haptics, and more ...
n Prof. Vendittelli (follow link at http://www.diag.uniroma1.it/vendittelli)
n Elective in Robotics (12 credits) or Control Problems in Robotics (6 credits)
n I-II semesters, starting this semester
n 4 modules of 3 credits (for CPR, students take 2 modules out of the 4 in EiR)
n research-related subjects: e.g., physical Human-Robot Interaction (pHRI)
n multiple instructors www.diag.uniroma1.it/vendittelli/EIR
n Probabilistic Robotics (6 credits), I semester, year 2
n Least Squares state estimation, Kalman filter, SLAM
n Prof. Grisetti sites.google.com/diag.uniroma1.it/probabilistic-robotics-2024-25
n Robot Programming (3 credits, no mark), I semester, year 1
n robot programming using C++, modules with ROS, embedded real-time coding
n Prof. Grisetti sites.google.com/diag.uniroma1.it/robot-programming-2024-25
Robotics 1 33
Robotics around the world...
Springer Handbook of Robotics (2nd Edition, July 2016)
Robotics 1 full video on https://vimeo.com/173394878 (free login) 34
I-RIM conference 2024
Gazometro Roma, 25-27 October 2024
n Institute of Robotics and Intelligent Machines (I-RIM)
n national association of academic and industrial stakeholders
n created in 2019
n check the web site for multiple initiatives: https://i-rim.it/en
n 6th national I-RIM conference
n scientific presentations, workshops, expo, demos …
n in conjunction with MakerFaire Europe
n registration for students is very low
n Student staff may be needed!
Robotics 1 35
