Algorithms for SensorBased Robotics Introduction and Background Computer

Algorithms for Sensor-Based Robotics: Introduction and Background Computer Science 436/636 https: //cirl. lcsr. jhu. edu/Sensor. Based. Robotics/ Greg Hager Simon Leonard 2/19/2021 CS 436/636, G. D. Hager

Admin https: //cirl. lcsr. jhu. edu/sensorbasedrobotics • Instructors: Greg Hager, Simon Leonard, Chris Paxton • Tuesday/Thursday: 15 h-16 h 15 • Lecture slides are posted on the web page and will be updated after each lecture – The current online slides are from last year • Office hours – Simon: By appointment – Greg: By appointment – Chris: TBD 2/19/2021 CS 436/636, G. D. Hager

Admin • Newsgroup: – Blackboard – Piazza • Last year both were used interchangeably (and that’s not a good thing) – So let’s decide which one: 2/19/2021 CS 436/636, G. D. Hager

Robotics at JHU • One of the top 10 universities for robotics in the US – Arguably the leader medical robotics • Laboratory for Computational Sensing and Robotics (LCSR) – Built in 2007 – Interdisciplinary: CS, ME, EE – 11 research labs (medical robotics, underwater robotics, drones, space robotics, …) 2/19/2021 CS 436/636, G. D. Hager

Robotics at JHU • Courses: – Robot Systems Programming (ME 530. 707) • Lots of ROS, lots of programming (Whitcomb) – Robot Devices, Kinematics, Dynamics, and Control (ME 530. 646) • Lots of mathematics modeling (Cowan) – Algorithms for Sensor-Based Robotics (CS 600. 436/636) (Hager, Leonard) • Algorithms: path planning, mapping, localization – Intro to Computer-Integrated Surgery (CS 600. 145) • Introduction to CIS – Computer-Integrated Surgery I (CS 600. 444/645) (Taylor) – Computer-Integrated Surgery II (CS 600. 445/646) (Taylor) • Projects 2/19/2021 CS 436/636, G. D. Hager

Robotics • Two broad categories (www. ifr. org) – Industrial Robots Bot. Junkie WIRED – Service Robots Discovering for Life Boston Dynamics

Industrial Robots Industrial robot as defined by ISO 8373: (www. ifr. org) “An automatically controlled, reprogrammable, multipurpose manipulator programmable in three or more axes, which may be either fixed in place or mobile for use in industrial automation applications. ” Reprogrammable: whose programmed motions or auxiliary functions may be changed without physical alterations; Multipurpose: capable of being adapted to a different application with physical alterations; Physical alterations: alteration of the mechanical structure or control system except for changes of programming cassettes, ROMs, etc. 2/19/2021 CS 436/636, G. D. Hager

Industrial Robots • 178, 132 units sold in 2013 (www. ifr. org) • Automotive, chemical, rubber & plastics, food industries • Total worldwide operational units at the end of 2013 was in the range of 1, 333, 000 to 1, 600, 000 • US$9. 5 billion in sales (market value for robot systems is US$29 billion). www. ifr. org

Service Robots • Pretty much everything else that is not an “industrial robot” ISO-8373: – A robot is an actuated mechanism programmable in two or more axes with a degree of autonomy, moving within its environment, to perform intended tasks. Autonomy in this context means the ability to perform intended tasks based on current state and sensing, without human intervention. – A service robot is a robot that performs useful tasks for humans or equipment excluding industrial automation application. Note: The classification of a robot into industrial robot or service robot is done according to its intended application. – Personal service robot is a service robot used for a non-commercial task. Examples are domestic servant robot, automated wheelchair, personal mobility assist robot, and pet exercising robot. – Professional service robot is a service robot used for a commercial task. Examples are cleaning robot for public places, delivery robot in offices.

Service Robots • 21, 000 professional service robots units sold in 2013 (www. ifr. org) • US$3. 57 in sales in 2013 • About 63, 500 units sold between 1996 and 2008, but about 100, 000 units were sold since. – Military (45% of service robots), drones (12%), ground vehicles (4%) – Farming: Milking robots (24%), mobile barn cleaner (Roomba for barns) and robotic fencer (4%) – Field robots (28%): mining, underwater, forest – Medical robots (6%) • Most valuable service robots with an average price of US$1. 5 million • 4 million personal service robot units sold in 2013 – Vacuum cleaners, lawn mowers – Assisted living (physical disabilities, elders)

This Course: Two Key Problems 1. 2. Motion Planning: Given a mechanism, a known environment, and a known start and goal, compute a set of inputs (e. g. a joint trajectory) that moves the robot from start to goal without collision. WIRED Localization and Mapping: Given no a priori knowledge, use information from sensing and motion to simultaneously compute a model of the environment (a map) and robot location within the environment. CNN

What is Motion Planning? • Determining where to go. . . more than a search (or it is a geometric search) 2/19/2021 CS 436/636, G. D. Hager

Basic Path Planning Problem Statement: Compute a continuous sequence of collision-free robot configurations 2 D EXAMPLE: connecting the initial and goal configurations Geometry of environment Geometry and kinematics of robot Initial and goal configurations Path Planner Collision-free path

Localization and Mapping • What is the world around me (mapping/geometry) – sense from various positions – integrate measurements to produce map – assumes perfect knowledge of position • Where am I in the world – – sense relate sensor readings to a world model compute location relative to model assumes a perfect world model • Together, these are SLAM (Simultaneous Localization and Mapping) – Can you localize yourself without a map? – Can you create a map if you cannot localize yourself? – Can you localize yourself and create a map at the same time?

Can we assume? • Perfect sensors? – What information – Uncertainty • Perfect control? – What controls? – Uncertainty • Perfect thinking? – Knowledge of the world? Complete? – Processing the world? Everything? 2/19/2021 CS 436/636, G. D. Hager

Probabilistic Methods/Drug Design Asada Nauyo robotics. standord. edu

Think About Dynamics Too! • Following a path and avoiding obstacles is one thing • Given other constraints (i. e. maximum acceleration, maximum torque), is there a path that maximizes ____? Scissors jump vs Fosbury flop Throwing a fast ball vs

Think About Dynamics, Too! Visually controlled juggling 3 R manipulator • 3 -5 full frames / bounce • used half-frames • estimated single-frame motion Al Rizzi, Noah Cowan, Rob Burridge, & Dan Koditschek 2/19/2021 • obstacle avoidance ? ! CS 436/636, G. D. Hager

SLAM • Suppose you wake up in the desert and you need to find your way around --- how do you do it? • A Chicken and Egg problem? 2/19/2021 CS 436/636, G. D. Hager

2/19/2021 CS 436/636, G. D. Hager

This Course: • We will focus on two robotics problems: – Robot Motion Planning: How do I get from A to B in an environment – Robot Localization and Mapping: What is the structure of space, and where am I in it. • • We will primarily focus on algorithms, their analysis, and their implementation (when possible) from real sensor data We will focus on the *static* planning problem – the world does not change – our own momentum does not matter • Most of our work will be in simulation although some of you will be able to apply it to real robots 2/19/2021 CS 436/636, G. D. Hager

Book http: //motionplanning. com Principles of Robot Motion: Theory, Algorithms, and Implementations H. Choset, K. M. Lynch, S. Hutchinson, G. Kantor, W. Burgard, L. E. Kavraki and S. Thrun, MIT Press, Boston, 2005. 2/19/2021 CS 436/636, G. D. Hager

Other Books Planning Algorithms Steven Lavalle, Cambridge University Prress, 2006 Robot Motion Planning, Jean-Claude Latombe, Kluwer, 1991. Free download: http: //planning. cs. uiuc. edu/ An Introduction to AI/Robotics Robin Murphy MIT Press, 2006 Probabilistic Robotics S. Thrun, W. Burgard, D. Fox MIT Press, 2006 2/19/2021 A Mathematical Introduction to Robotic Manipulation Murray, Li, Sastry CS 436/636, G. D. Hager

This Course: • Text: Principles of Robot Motion: Theory, Algorithms, and Implementation, Choset et. al, 2005 • • • Chap 1: Introduction Chap 3: Kinematics and Configuration Space Chap 4: Potential Field Methods Collision Detection Chap 5: Roadmaps Chap 6: Cell decompositions EXAM Chap 7: Sampling-Based Motion Planning Chap 8: Kalman-Filter based Localization Chap 9: Bayesian Methods for Localization EXAM 2/19/2021 CS 436/636, G. D. Hager

Assignments – First assignment is paper/pencil – Others are coding • Using Robot Operating System (ROS) • You will need access to a Linux computer with Ubuntu (dual boot, virtual machine, etc. ) – Some dual boot computers are available in Wyman 157 – Working remotely will not work well given that the assignments require graphical software • If you have a laptop/desktop I highly recommend installing Linux on it – We will host a session on how to do this if you need help – We will do the best support we can but if it doesn’t work you will need to use the computers in 157 • ROS packages will be provided and you will edit lines of code and fill in placeholders (methods, functions, etc. ) 2/19/2021 CS 436/636, G. D. Hager

Move. It! is a software for mobile manipulation, incorporating the latest advances in motion planning, manipulation, 3 D perception, kinematics, control and navigation. http: //moveit. ros. org/ 2/19/2021 CS 436/636, G. D. Hager