View Chapter

Chapter 74 — Learning from Humans

Aude G. Billard, Sylvain Calinon and Rüdiger Dillmann

This chapter surveys the main approaches developed to date to endow robots with the ability to learn from human guidance. The field is best known as robot programming by demonstration, robot learning from/by demonstration, apprenticeship learning and imitation learning. We start with a brief historical overview of the field. We then summarize the various approaches taken to solve four main questions: when, what, who and when to imitate. We emphasize the importance of choosing well the interface and the channels used to convey the demonstrations, with an eye on interfaces providing force control and force feedback. We then review algorithmic approaches to model skills individually and as a compound and algorithms that combine learning from human guidance with reinforcement learning. We close with a look on the use of language to guide teaching and a list of open issues.

Learning compliant motion from human demonstration

Author  Aude Billard

Video ID : 478

This video illustrates how one can teach a robot to display the right amount of stiffness to perform a task successfully. Decrease in stiffness is demonstrated by shaking the robot, while increase in stiffness is conveyed by pressing on the robot's arm (pressure being measured through tactile sensors along the robot's arm). Reference: K. Kronander,A. Billard: Learning compliant manipulation through kinesthetic and tactile human-robot interaction, IEEE Trans. Haptics 7(3), 367-380 (2013); doi: 10.1109/TOH.2013.54 .

Chapter 21 — Actuators for Soft Robotics

Alin Albu-Schäffer and Antonio Bicchi

Although we do not know as yet how robots of the future will look like exactly, most of us are sure that they will not resemble the heavy, bulky, rigid machines dangerously moving around in old fashioned industrial automation. There is a growing consensus, in the research community as well as in expectations from the public, that robots of the next generation will be physically compliant and adaptable machines, closely interacting with humans and moving safely, smoothly and efficiently - in other terms, robots will be soft.

This chapter discusses the design, modeling and control of actuators for the new generation of soft robots, which can replace conventional actuators in applications where rigidity is not the first and foremost concern in performance. The chapter focuses on the technology, modeling, and control of lumped parameters of soft robotics, that is, systems of discrete, interconnected, and compliant elements. Distributed parameters, snakelike and continuum soft robotics, are presented in Chap. 20, while Chap. 23 discusses in detail the biomimetic motivations that are often behind soft robotics.

DLR Hand Arm System throwing a ball and Justin catching it

Author  Alin Albu-Schäffer, Thomas Bahls, Berthold Bäuml, Maxime Chalon, Markus Grebenstein, Oliver Eiberger, Werner Friedl, Hannes Höppner, Dominic Lakatos, Nico Mansfeld, Florian Petit, Jens Reinecke, Roman Weitschat, Sebastian Wolf, Tilo Wüsthoff

Video ID : 547

The DLR Hand Arm System throws a ball and Justin catches it. There is no data connection between the two systems. Justin catches the ball by visual observation.

Chapter 69 — Physical Human-Robot Interaction

Sami Haddadin and Elizabeth Croft

Over the last two decades, the foundations for physical human–robot interaction (pHRI) have evolved from successful developments in mechatronics, control, and planning, leading toward safer lightweight robot designs and interaction control schemes that advance beyond the current capacities of existing high-payload and highprecision position-controlled industrial robots. Based on their ability to sense physical interaction, render compliant behavior along the robot structure, plan motions that respect human preferences, and generate interaction plans for collaboration and coaction with humans, these novel robots have opened up novel and unforeseen application domains, and have advanced the field of human safety in robotics.

This chapter gives an overview on the state of the art in pHRI as of the date of publication. First, the advances in human safety are outlined, addressing topics in human injury analysis in robotics and safety standards for pHRI. Then, the foundations of human-friendly robot design, including the development of lightweight and intrinsically flexible force/torque-controlled machines together with the required perception abilities for interaction are introduced. Subsequently, motionplanning techniques for human environments, including the domains of biomechanically safe, risk-metric-based, human-aware planning are covered. Finally, the rather recent problem of interaction planning is summarized, including the issues of collaborative action planning, the definition of the interaction planning problem, and an introduction to robot reflexes and reactive control architecture for pHRI.

Justin: A humanoid upper body system for two-handed manipulation experiments

Author  Christoph Borst, Christian Ott, Thomas Wimböck, Bernhard Brunner, Franziska Zacharias, Berthold Bäuml

Video ID : 626

This video presents a humanoid two-arm system developed as a research platform for studying dexterous two-handed manipulation. The system is based on the modular DLR-Lightweight-Robot-III and the DLR-Hand-II. Two arms and hands are combined with a 3-DOF movable torso and a visual system to form a complete humanoid upper body. The diversity of the system is demonstrated by showing the mechanical design, several control concepts, the application of rapid prototyping and hardware-in-the-loop (HIL) development, as well as two-handed manipulation experiments and the integration of path planning capabilities.

Chapter 17 — Limbed Systems

Shuuji Kajita and Christian Ott

A limbed system is a mobile robot with a body, legs and arms. First, its general design process is discussed in Sect. 17.1. Then we consider issues of conceptual design and observe designs of various existing robots in Sect. 17.2. As an example in detail, the design of a humanoid robot HRP-4C is shown in Sect. 17.3. To design a limbed system of good performance, it is important to take into account of actuation and control, like gravity compensation, limit cycle dynamics, template models, and backdrivable actuation. These are discussed in Sect. 17.4.

In Sect. 17.5, we overview divergence of limbed systems. We see odd legged walkers, leg–wheel hybrid robots, leg–arm hybrid robots, tethered walking robots, and wall-climbing robots. To compare limbed systems of different configurations,we can use performance indices such as the gait sensitivity norm, the Froude number, and the specific resistance, etc., which are introduced in Sect. 17.6.

Hexapod ParaWalker-II

Author  Yusuke Ota

Video ID : 520

A twin-frame walking robot, which is a reduced-DOF practical walking robot, developed by Dr. Ota and Prof. Hirose.

Chapter 21 — Actuators for Soft Robotics

Alin Albu-Schäffer and Antonio Bicchi

Although we do not know as yet how robots of the future will look like exactly, most of us are sure that they will not resemble the heavy, bulky, rigid machines dangerously moving around in old fashioned industrial automation. There is a growing consensus, in the research community as well as in expectations from the public, that robots of the next generation will be physically compliant and adaptable machines, closely interacting with humans and moving safely, smoothly and efficiently - in other terms, robots will be soft.

This chapter discusses the design, modeling and control of actuators for the new generation of soft robots, which can replace conventional actuators in applications where rigidity is not the first and foremost concern in performance. The chapter focuses on the technology, modeling, and control of lumped parameters of soft robotics, that is, systems of discrete, interconnected, and compliant elements. Distributed parameters, snakelike and continuum soft robotics, are presented in Chap. 20, while Chap. 23 discusses in detail the biomimetic motivations that are often behind soft robotics.

VSA-Cube: Arm with high and low stiffness preset

Author  Centro di Ricerca "E. Piaggio"

Video ID : 470

A modular 2-DOF arm, built with VSA-cube actuation units, performing high- and low-stiffness behaviors.

Chapter 11 — Robots with Flexible Elements

Alessandro De Luca and Wayne J. Book

Design issues, dynamic modeling, trajectory planning, and feedback control problems are presented for robot manipulators having components with mechanical flexibility, either concentrated at the joints or distributed along the links. The chapter is divided accordingly into two main parts. Similarities or differences between the two types of flexibility are pointed out wherever appropriate.

For robots with flexible joints, the dynamic model is derived in detail by following a Lagrangian approach and possible simplified versions are discussed. The problem of computing the nominal torques that produce a desired robot motion is then solved. Regulation and trajectory tracking tasks are addressed by means of linear and nonlinear feedback control designs.

For robots with flexible links, relevant factors that lead to the consideration of distributed flexibility are analyzed. Dynamic models are presented, based on the treatment of flexibility through lumped elements, transfer matrices, or assumed modes. Several specific issues are then highlighted, including the selection of sensors, the model order used for control design, and the generation of effective commands that reduce or eliminate residual vibrations in rest-to-rest maneuvers. Feedback control alternatives are finally discussed.

In each of the two parts of this chapter, a section is devoted to the illustration of the original references and to further readings on the subject.

Inverse dynamics control for a flexible link

Author  Wayne Book

Video ID : 778

A single flexible link with rotation at its base is controlled by computing the stable inverse dynamics of the flexible system associated with the desired trajectory for the end-effector. This feedforward command is made more robust by the addition of a suitable PD feedback control at the joint. Because of the non-minimum phase nature of the tip output, the resulting input command is non-causal, starting ahead of the actual output trajectory (pre-shaping the link) and ending after (discharging the link). Comparison is made with a PD joint control using a step reference input and with a full state feedback (utilizing strain gauge signals and their rates) and a nominal trajectory command. The inverse dynamics control demonstrates superiority both in terms of overshoot and residual vibrations. References: 1. D.-S. Kwon: An Inverse Dynamic Tracking Control for a Bracing Flexible Manipulator, Dissertation, School of Mechanical Engineering, Georgia Institute of Technology, Atlanta, (1991); 2. D.-S. Kwon, W.J. Book: A time-domain inverse dynamic tracking control of a single-link flexible manipulator, ASME J. Dyn. Syst. Meas. Control 116, 193-200 (1994); doi: 10.1115/1.2899210

Chapter 14 — AI Reasoning Methods for Robotics

Michael Beetz, Raja Chatila, Joachim Hertzberg and Federico Pecora

Artificial intelligence (AI) reasoning technology involving, e.g., inference, planning, and learning, has a track record with a healthy number of successful applications. So can it be used as a toolbox of methods for autonomous mobile robots? Not necessarily, as reasoning on a mobile robot about its dynamic, partially known environment may differ substantially from that in knowledge-based pure software systems, where most of the named successes have been registered. Moreover, recent knowledge about the robot’s environment cannot be given a priori, but needs to be updated from sensor data, involving challenging problems of symbol grounding and knowledge base change. This chapter sketches the main roboticsrelevant topics of symbol-based AI reasoning. Basic methods of knowledge representation and inference are described in general, covering both logicand probability-based approaches. The chapter first gives a motivation by example, to what extent symbolic reasoning has the potential of helping robots perform in the first place. Then (Sect. 14.2), we sketch the landscape of representation languages available for the endeavor. After that (Sect. 14.3), we present approaches and results for several types of practical, robotics-related reasoning tasks, with an emphasis on temporal and spatial reasoning. Plan-based robot control is described in some more detail in Sect. 14.4. Section 14.5 concludes.

RoboEarth final demonstrator

Author  Gajamohan Mohanarajah

Video ID : 706

This video made in 2014 summarizes the final demonstrator of the joint project RoboEarth -- A World Wide Web for robots (http://roboearth.org/). The demonstrator includes four robots collaboratively working together to help patients in a hospital. These robots used their common knowledge base and infrastructure in the following ways: 1. a knowledge repository to share and learn from each others' experience, 2. a communication medium to perform collaborative tasks, and 3. a computational resource to offload some of their heavy computational load.

Chapter 62 — Intelligent Vehicles

Alberto Broggi, Alex Zelinsky, Ümit Özgüner and Christian Laugier

This chapter describes the emerging robotics application field of intelligent vehicles – motor vehicles that have autonomous functions and capabilities. The chapter is organized as follows. Section 62.1 provides a motivation for why the development of intelligent vehicles is important, a brief history of the field, and the potential benefits of the technology. Section 62.2 describes the technologies that enable intelligent vehicles to sense vehicle, environment, and driver state, work with digital maps and satellite navigation, and communicate with intelligent transportation infrastructure. Section 62.3 describes the challenges and solutions associated with road scene understanding – a key capability for all intelligent vehicles. Section 62.4 describes advanced driver assistance systems, which use the robotics and sensing technologies described earlier to create new safety and convenience systems for motor vehicles, such as collision avoidance, lane keeping, and parking assistance. Section 62.5 describes driver monitoring technologies that are being developed to mitigate driver fatigue, inattention, and impairment. Section 62.6 describes fully autonomous intelligent vehicles systems that have been developed and deployed. The chapter is concluded in Sect. 62.7 with a discussion of future prospects, while Sect. 62.8 provides references to further reading and additional resources.

Pedestrian detection

Author  Alberto Broggi, Alexander Zelinsky, Ümit Ozgüner, Christian Laugier

Video ID : 839

This video demonstrates pedestrian detection using stereo vision to achieve robustness.

Chapter 58 — Robotics in Hazardous Applications

James Trevelyan, William R. Hamel and Sung-Chul Kang

Robotics researchers have worked hard to realize a long-awaited vision: machines that can eliminate the need for people to work in hazardous environments. Chapter 60 is framed by the vision of disaster response: search and rescue robots carrying people from burning buildings or tunneling through collapsed rock falls to reach trapped miners. In this chapter we review tangible progress towards robots that perform routine work in places too dangerous for humans. Researchers still have many challenges ahead of them but there has been remarkable progress in some areas. Hazardous environments present special challenges for the accomplishment of desired tasks depending on the nature and magnitude of the hazards. Hazards may be present in the form of radiation, toxic contamination, falling objects or potential explosions. Technology that specialized engineering companies can develop and sell without active help from researchers marks the frontier of commercial feasibility. Just inside this border lie teleoperated robots for explosive ordnance disposal (EOD) and for underwater engineering work. Even with the typical tenfold disadvantage in manipulation performance imposed by the limits of today’s telepresence and teleoperation technology, in terms of human dexterity and speed, robots often can offer a more cost-effective solution. However, most routine applications in hazardous environments still lie far beyond the feasibility frontier. Fire fighting, remediating nuclear contamination, reactor decommissioning, tunneling, underwater engineering, underground mining and clearance of landmines and unexploded ordnance still present many unsolved problems.

UNMACA: Demining Afghanistan

Author  James P. Trevelyan

Video ID : 571

This is a high-quality video made partly with the aim of seeking funds to help complete demining projects in Afghanistan. This video has been included because researchers can see plenty of examples of realistic field conditions under which demining is being done in Afghanistan. It is essential for researchers to have an accurate appreciation of the real field conditions before considering expensive research projects. There are plenty of opportunities to see manual mine clearance. Current-generation demining machines don't work here because of the very hard and rocky ground. There is an interesting segment showing the Bamyan site. The sentiments expressed by deminers are genuine, in my experience. I have met many similarly dedicated Afghan deminers, and they are selected for their dedication, attitude to nation-building, courage, and conscientious work ethic. They are justly proud of the work they do, and their uniforms and equipment set them apart from most other Afghans and give them a real sense of respect. Note that winter rains and summer storms wash mud over mines, encasing them in what later turns to hard, cement-like soil. It is hard physical work demanding sensitive hands, care, and attention to detail. For more information see: http://school.mech.uwa.edu.au/~jamest/demining/countries/afghan/minefields-afghan.html

Chapter 41 — Active Manipulation for Perception

Anna Petrovskaya and Kaijen Hsiao

This chapter covers perceptual methods in which manipulation is an integral part of perception. These methods face special challenges due to data sparsity and high costs of sensing actions. However, they can also succeed where other perceptual methods fail, for example, in poor-visibility conditions or for learning the physical properties of a scene.

The chapter focuses on specialized methods that have been developed for object localization, inference, planning, recognition, and modeling in activemanipulation approaches.We concludewith a discussion of real-life applications and directions for future research.

Modeling articulated objects using active manipulation

Author  Juergen Strum

Video ID : 78

The video illustrates a mobile, manipulation robot that interacts with various articulated objects, such as a fridge and a dishwasher, in a kitchen environment. During interaction, the robot learns their kinematic properties such as the rotation axis and the configuration space. Knowing the kinematic model of these objects improves the performance of the robot and enables motion planning. Service robots operating in domestic environments are typically faced with a variety of objects they have to deal with to fulfill their tasks. Some of these objects are articulated such as cabinet doors and drawers, or room and garage doors. The ability to deal with such articulated objects is relevant for service robots, as, for example, they need to open doors when navigating between rooms and to open cabinets to pick up objects in fetch-and-carry applications. We developed a complete probabilistic framework that enables robots to learn the kinematic models of articulated objects from observations of their motion. We combine parametric and nonparametric models consistently and utilize the advantages of both methods. As a result of our approach, a robot can robustly operate articulated objects in unstructured environments. All software is available open-source (including documentation and tutorials) on http://www.ros.org/wiki/articulation.