Advertisment

Session 3: Industrial Perception and Autonomy

No performance without perception—smart machine vision as key technology of modern robotics

What good is the smartest machine if it can't utilize its potential? A robot without high-end environmental recognition is like a race car without steering wheel. As autonomous interaction between machines becomes increasingly important, it must be supported by modern sensor technology. What technologies will be available to implement this in the future? And what deficits remain? How will the evolution of machine vision affect the development of robotics in the coming years?

Four scientists will bring their audience up to speed on perception and autonomy.

Dr.-Ing. Gunther Kegel, Pepperl+Fuchs
Dr. Alfred Rizzi, Boston Dynamics
Prof. Matthew Mason, Berkshire Grey
Dr. Christoph Peylo, Robert Bosch
Session Chair
Advertisment

Robotics could already have advanced by several decades more, claims Dr. Gunther Kegel, CEO of Pepperl+Fuchs. This has been prevented by the performance deficit of industrial sensor and machine vision technology. While computing power and storage capacity evolved continuously, environment recognition has fallen by the wayside for a long time. Gunther Kegel outlines the status quo and describes both visions and requirements for machine vision applications of the future.

According to Dr. Alfred Rizzi and his research on legged robots, the future uses limbs, not wheels. The Chief Engineer at Boston Dynamics specializes in machines that mimic the gait of humans or various animal species. Unlike their rolling predecessors, these machines can conquer parts of the world previously inaccessible to robots.

Prof. Matthew Mason, Chief Engineer at Berkshire Grey, the leading U.S. technology company for AI in retail, takes a look at the future of robot-based automation in intralogistics.

Trust is not a digital concept. At least not yet, according to Dr. Christoph Peylo of Robert Bosch GmbH. Our interpersonal communication has an established understanding of concepts like ‘trust’ or ‘believe’. In the digital world, however, that is not the case. In his presentation, Christoph Peylo will explain how artificial intelligence can help set standards for a ‘digital trust concept’.

Speakers and presentations in this session

“The importance of sensors and cognitive capabilities for next generation robotic AI applications”

The lack of cognitive capabilities and the performance of industrial sensors have been the bottleneck for robot applications in and outside of the industry since decades. Whereas computational power and memory capacity have grown substantially, and motors and drivers have turned bulky robots into high precision fast moving light weight kinematic machines, the cognitive capabilities to understand and interpret the robot’s environment have not been growing with the same pace. But data of physical, geometrical and even mental status of the environment are essential to feed the well-known and powerful AI algorithms in applications outside of the well-defined industrial world. The presentation will document the state of the art and gives some trends, visions and requirements for the next generation of sensors that will help to increase cognitive capabilities to a further level.

The importance of sensors and cognitive capabilities for next generation robotic AI applications

The lack of cognitive capabilities and the performance of industrial sensors have been the bottleneck for robot applications in and outside of the industry since decades. Whereas computational power and memory capacity have grown substantially and motors and drivers have turned bulky robots into high precision fast moving light weight kinematic machines, the cognitive capabilities to understand and interpret the robot’s environment have not been growing with the same pace. But data of physical, geometrical and even mental status of the environment are essential to feed the well-known and powerful AI algorithms in applications outside of the well-defined industrial world. The presentation will document the state of the art and gives some trends, visions and requirements for the next generation of sensors that will help to increase cognitive capabilities to a further level.

Hightech-Summit Session 3: Industrial Perception and Autonomy

„Developing and Deploying Capable Legged Mobile Manipulation Robots”

At Boston Dynamics over the past decade, we have made significant progress in developing locomotion capability that rivals humans and animals. One important feature of these systems is that they are easy for an operator to drive, giving users the ability to access a significant fraction of the world that was previously inaccessible via wheeled and tracked robots. As we begin to deliver legged mobile robots with arms and grippers, we seek to do the same thing in the mobile manipulation space and provide reactive control strategies that allow users/operators and higher-level control systems to effectively use a robot to reliably interact with its environment to pick, place, grasp, and manipulate useful objects.

Dr. Alfred Rizzi is currently the Chief Scientist at Boston Dynamics where he directs research and product development based on novel locomotion and mobile manipulation systems. Prior to joining Boston Dynamics, in 2006, he was an Associate Research Professor in the Robotics Institute at Carnegie Mellon University where he directed research projects focused on hybrid sensor-based control of complex and distributed dynamical systems. Dr. Rizzi received the Sc.B, degree in electrical engineering from the Massachusetts Institute of Technology in 1986. He received the M.S. and Ph.D. from Yale University in 1990 and 1994 respectively.

Developing and Deploying Capable Legged Mobile Manipulation Robots

At Boston Dynamics over the past decade we have made significant progress in developing locomotion capability that rivals humans and animals. One important feature of these systems is that they are easy for an operator to drive, giving users the ability to access a significant fraction of the world that was previously inaccessible via wheeled and tracked robots. As we begin to deliver legged mobile robots with arms and grippers we seek to do the same thing in the mobile manipulation space and provide reactive control strategies that allow users/operators and higher level control systems to effectively use a robot to reliably interact with its environment to pick, place, grasp, and manipulate useful objects.

Hightech-Summit Session 3: Industrial Perception and Autonomy

“Robotics and Warehouse Automation: from robotics research to Berkshire Grey”

This talk tells the Berkshire Grey story, how it adapted robotics technology from research labs to transform warehouses and distribution centers, leading to a public listing on NASDAQ – the first since iRobot 15 years ago. Berkshire Grey produces automated systems for e-commerce order fulfillment, parcel sortation, store replenishment, and related operations in warehouses, distribution centers, and in the back ends of stores. The talk will discuss the fundamental issues in warehouse automation, the technology and solutions developed by Berkshire Grey, and future challenges.

Matthew T. Mason is the Chief Scientist at Berkshire Grey. He earned the BS, MS, and PhD degrees at the MIT AI Lab. At Carnegie Mellon University he served ten years as Chair of the Robotics PhD program, and another ten years as Director of the CMU Robotics Institute. He is a Fellow of the AAAI, the IEEE, and the ACM. He is a winner of the System Development Foundation Prize, the IEEE RAS Pioneer Award, and the 2018 IEEE Technical Field Award in Robotics and Automation.

Robotics and Warehouse Automation: from robotics research to Berkshire Grey

This talk tells the Berkshire Grey story, how it adapted robotics technology from research labs to transform warehouses and distribution centers, leading to a public listing on NASDAQ -- the first since iRobot 15 years ago. Berkshire Grey produces automated systems for e-commerce order fulfillment, parcel sortation, store replenishment, and related operations in warehouses, distribution centers, and in the back ends of stores. The talk will discuss the fundamental issues in warehouse automation, the technology and solutions developed by Berkshire Grey, and future challenges.

Hightech-Summit Session 3: Industrial Perception and Autonomy

“Trustworthy AI and Digital Trust: Trust lies in the eye of the beholder”

The physical world offers a sufficiently stable and agreed on basis for most human interactions, since human beings, by and large, share a similar set of mechanisms to perceive the world and interact with and communicate about it. Therefore, the underlying ontological commitments and assumptions of statements like “trust”, “believe” etc., can, most of the time, be taken for granted and do not have to be explained explicitly.

This, however, is not the case within the digital world. What commonly is referred to as “digital world”, is only loosely defined by agreed on protocols on a communication network, e.g., the internet, and tools, services, and applications to communicate and interact with networked devices. Protocols, tools and underlying networks are under constant development and change, so the digital world is an extremely brittle environment.

Thus, “digital trust” will not work as a simple mapping of the relevant dimensions of “trust” to their respective digital representations in an existing digital world, since there is no stable digital world. And AI brings in even more volatility. The respective domain and scope of “trust” and the relationship of the involved entities must be defined and created for any context in which trustworthy digital interactions should take place. Thus, “Digital Trust” specifies the relationship between two entities with respect to a specific task that requires entitlement, competence and adherence to certain pre-defined standards, protocols or principles. For keeping track of the expectations, learning from experience and adapt to specific behavior patterns, AI is an extremely powerful technology. Thus, AI can support to establish Digital Trust by adapting a digital product or service to the individual expectations of people during the whole life cycle of a product.

Experiencing “Digital Trust” will be the corresponding digital counterpart of today’s core quality and value proposition of the non-digital world.

Dr. Christoph Peylo leads the project “Digital Trust” in Bosch’s CDO organization. Before that he established and lead Bosch’s Center for Artificial Intelligence (BCAI). He was member of the high-level expert group on AI of the European Commission. Prior to joining Robert Bosch GmbH in 2017 he worked at Deutsche Telekom Laboratories in Berlin in the area of AI, Cybersecurity, Industrie 4.0, and IoT.

Before joining Deutsche Telekom in 2006 he worked in various positions from software engineer to managing director of a software company.

Christoph Peylo has studied Computer Science, Computational Linguistics and Artificial Intelligence and acquired his Ph.D. at the University of Osnabrück in the field of AI.

Trustworthy AI and Digital Trust: Trust lies in the eye of the beholder

The physical world offers a sufficiently stable and agreed on basis for most human interactions, since human beings, by and large, share a similar set of mechanisms to perceive the world and interact with and communicate about it. Therefore, the underlying ontological commitments and assumptions of statements like “trust”, “believe” etc., can, most of the time, be taken for granted and do not have to be explained explicitly.

This, however, is not the case within the digital world. What commonly is referred to as “digital world”, is only loosely defined by agreed on protocols on a communication network, e.g. the Internet, and tools, services, and applications to communicate and interact with networked devices. Protocols, tools and underlying networks are under constant development and change, so the digital world is an extremely brittle environment.

Thus, “digital trust” will not work as a simple mapping of the relevant dimensions of “trust” to their respective digital representations in an existing digital world, since there is no stable digital world. And AI brings in even more volatility. The respective domain and scope of “trust” and the relationship of the involved entities have to be defined and created for any context in which trustworthy digital interactions should take place. Thus, “Digital Trust” specifies the relationship between two entities with respect to a specific task that requires entitlement, competence and adherence to certain pre-defined standards, protocols or principles. For keeping track of the expectations, learning from experience and adapt to specific behavior patterns, AI is an extremely powerful technology. Thus, AI can support to establish Digital Trust by adapting a digital product or service to the individual expectations of people during the whole life cycle of a product.

Experiencing “Digital Trust” will be the corresponding digital counterpart of today’s core quality and value proposition of the non-digital world.

Hightech-Summit Session 3: Industrial Perception and Autonomy

Session Chair

Industrial Perception and Autonomy will be hosted by Session Chair Prof. Dr. rer. nat. Edward G. Krubasik, Honorary Professor at the TUM School of Management (TU Munich), Chairman of the Industry Advisory Board at the Munich Institute of Robotics and Machine Intelligence (MIRMI).


Technologies have the potential to assist humans and improve our quality of life. However, it is essential to design technologies with a focus on the benefits for the many, and not the few. Question of social justice and equity need to move to the center of technology development, particularly in fields such as AI and robotics. To this end, we must foster interdisciplinary collaborations between the social sciences and AI research and address social, ethical and political questions in an integrated way already during technology development.

Advertisment
Advertisment
Advertisment
Advertisment
Advertisment
munich_i
Session 4: Learning, Understanding and Interaction

Human-machine communication: from autonomous driving to social interaction

Mehr erfahren
munich_i
Session 1: Next Generation Robotics

What will future robot generations look like, what tasks will they perform, and how will they differ from the robots we know today?

Mehr erfahren
munich_i
Session 2: Networking Intelligence

Future robots will be integrated into smart and self-learning systems – in industrial applications, medical engineering and pharmaceutical research.

Mehr erfahren