Meet Baxter. Baxter is a cobot, a collaborative robot. He is a new breed of robot. He can work in a highly unstructured environment, has more degrees of freedom (more joints), can apply logic to make decisions, and can communicate with other systems. He is a resident of the Accenture Innovation Lab in Cluj, but he is not from the future. He is already working in several factories around the world.
Baxter is part of a new wave of robots that will share our future – a card-carrying member of Industry 4.0, aka the fourth industrial revolution, where interconnected cyber-physical systems are taking over from mechatronics. But Baxter will not be alone: an open source Robot Operating System (ROS) is coming along nicely and it will bring high-level ROS functionality to turbocharge the low-level reliability and safety of industrial robot controllers – increasing programmability and control to turn limited capability Industry 2.0 and 3.0 industrial robots into more responsive and integrated members of a smart environment.
The majority of the organizations do not yet grasp the implications of the Industrial Internet (of Things) on their industries or their businesses. IoT provides the tools for the interconnection of cyber and physical systems, while robotics is advancing to provide the “smart workers” for the Smart Factory of the 4.0 world. Together they will bring profound change to the entire industrial ecosystem.
Imagine, for a moment, a world where sensors and machines “talk” to other machines and to humans, automating decisions and production in real time, based on environmental conditions, demand, market pricing, value chain performance or the performance of plant equipment and physical processes – all in real time, with decentralised decision-making adding intelligence all along the value chain.
While Industry 4.0 is not yet fully defined, it is going to deliver a full paradigm shift. The first Industrial Revolution was defined by the introduction of mechanical production facilities powered by water and steam. The second introduced mass production using electrical energy, and the third the integration of electronics, mechanics and information technology (the so-called mechatronic systems that automate production, aka the digital revolution). Industry 4.0 will feature the inter-connected cyber-physical systems.
Industry 4.0 will provide a set of principles that supports companies in implementing a Smart Factory vision, something that will impact value chains, workforces and business models … and turn entire industries on their heads.
In Industry 4.0, new value propositions will emerge and new value-based outcomes will be built on the new digital platforms and platform ecosystems. These offerings will be created by multiple, disparate companies from within and beyond traditional value chains, companies that collaborate to create a new ecosystem that delivers experiential ‘solutions’ to customers, companies that go far beyond the delivery of a single product.
In the Smart Factory of Industry 4.0, emerging technologies like IoT, Big Data and analytics, 3D Printing, Augmented Reality and Autonomous Robots will be first-class citizens.
The initial drivers for the adoption of IoT and Industry 4.0 technologies will be cost reduction and revenue growth. These can be achieved through the optimisation of asset usage, predictive maintenance and remote management, and through the development of alternative revenue streams through new products and services. However, increasing workforce productivity, safety and working conditions will also be an increasingly important driving force. Drones that inspect remote field equipment and operations, and solutions that minimize human exposure to harsh or dangerous conditions (e.g., chemicals and gases) are key examples. Robotics is clearly going to play a critical role here.
Robotics was first defined by the science fiction writer Isaac Asimov who also created the often referenced Three Laws of Robotics. Are we about to embark on a journey to a future were the ubiquity of intelligent robots make these laws serious security requirements rather than philosophical fare? We are further along the track than many imagine.
Forty years after their invention, there are about 10 million industrial robots working diligently around the world. This current generation of industrial robots, often called manipulators, basically consist of an arm with several joints and an end-effector (mostly a gripper) with a fixed or mobile base. They are used in car production, packaging and electronics. Automated Guided Vehicles with different navigation capabilities are also used to transport goods around in environments such as warehouses, ports or hospitals. Japan, in particular, makes heavy use of robots in its manufacturing industry.
In the next decade, an inflection point will be reached in many industries where deploying industrial robots will become more commercially viable due to decreasing hardware and software prices and increasing performance. The development of a standard computer operating system designed primarily for robots – ROS – will help trigger broad adoption.
It has only been three years since ROS became open source, but it is already causing quite a stir.
The Robot Operating System (ROS) is an open-source set of programs initially developed by Willow Garage and handed over to the Open Source Robotics Foundation in 2013. Among the main contributors to the development of ROS are Stanford University, Massachusetts Institute of Technology and the Technical University of Munich, Germany.
What is in the box? ROS offers an OS-like functionality in a heterogeneous distributed system. The OS-like services include hardware abstraction, low-level device control, implementation of commonly used functionality, message-passing between processes, and package management. The ROS runtime "graph" is a peer-to-peer network of processes (potentially distributed across machines) that are loosely coupled using the ROS communication infrastructure.
The ROS implements several different styles of communication, including synchronous remote procedure call (RPC)-style communication over services, asynchronous streaming of data over topics, and storage of data on a Parameter Server. This architecture makes it easy to distribute processing load among devices and allows ROS robots to communicate with each other as members of the same peer-to-peer network. It is also easy to switch nodes, allowing for simulation.
Although ROS is not a real-time operating system, it provides reasonable performance and, therefore, it is considered a “low-latency” framework.
When the ROS system loads, it reads the robot description parameters and initializes the runtime graph accordingly. Implementing ROS for a specific robot encompasses providing the specific parameters such as joint limits and arm segment lengths in a common description format (URDF) and implementing drivers for the specific hardware. Many robot platforms, ranging from two-wheeled mobile platforms to humanoids, and from hobby kits to industrial robots, run ROS or provide a ROS interface.
The ROS tools are language agnostic and therefore many languages can be used for ROS development. There are client libraries for Python, C++, LISP, Java, JavaScript and MATLAB. The supported OS platform is Ubuntu, with various ports to other Linux flavors.
ROS-Industrial Consortium started with two branches in the Americas and Europe in 2012 and 2013, respectively, to support standards and create ROS libraries, tools and drivers for the industrial hardware. It aims at creating a community of industrial professionals and at developing robust and reliable software that meets the needs of industrial applications.
The value? By combining the relative strength of ROS with the existing technologies, the high-level ROS functionality can be applied to address the low-level reliability and safety of industrial robot controllers. The use of standard interfaces and architecture stimulates hardware-agnostic software development and reusability, and facilitates the integration of research results into industrial applications.
These advances in hardware and software are seeing the emergence of a new breed of robots – the collaborative robots or cobots.
The old-generation industrial robot is rigid. It can move at fixed speeds, is suitable for repetitive, high-volume operations and for handling uniform sized objects, and cannot make any decisions on its own.
In contrast, a modern cobot, like Baxter, can work in a highly unstructured environment, can execute more complex movements, apply logic to decision-making, and can communicate with other systems. Built with human collaboration in mind, cobots can easily be trained to do the tasks of unskilled workers. They also have more safety features.
Baxter is the creation of Rethink Robotics. Its founder, Rodney Brooks, is the co-founder and former chief technology officer of iRobot, makers of the popular robo-vacuum the Roomba, not to mention some scary military robots. Like iRobot, Rethink Robotics is making actual, for-sale robots, as opposed to research and PR-driven products like Honda's ASIMO. The company even sells accessories and extended warranties.
Baxter is a worker. It can pack things in boxes. It can inspect, sort, and align parts. Baxter can even do "light" assembly. Baxter has two arms, each with seven degrees of freedom (DOF), allowing it to move items from one location to another. The torso contains the computer and a few vacuum connections for the optional suction cup hands. The head is a tablet used to work Baxter's user interface (UI), to show what Baxter is currently paying attention to, and to communicate its status.
One of the differentiating features of Baxter is that unskilled personnel can use it, too. The robot can be trained to do a task without special knowledge – for example, by grabbing it by its wrist, and just performing the intended movements. There are navigation buttons and a knob on each arm and side of the torso for easy input and interaction.
While the screen would have you believe Baxter's eyes are on its face, they are really in its hands. Right next to the hand mount point, there is an array of sensors that are always pointed at the item Baxter is picking up or holding, so the robot will know if it runs out of items or if it drops something. It can check distance from objects by using the infrared depth sensor.
The head camera and the ultrasonic sensor array mounted above extend the perception capabilities. The end-effectors can be parallel grippers, suction cups (in single or multiple setup) or even third-party hardware and software accessories, such as general-purpose grippers resembling a human hand.
Baxter was built from the ground up with safety in mind. If its arm feels resistance from hitting an object (like a human), it will stop and resume its action if possible. Its ergonomic design prevents the fingers from being pinched. In addition, there is an emergency stop.
The torso hides a PC class computing power (3rd Gen Intel Core i7) running ROS.
While some of the Industry 4.0 vision will only be achieved in a couple of years, with its impact only perceived after a decade, we cannot say the same about cobots – they are already here!
A World Economic Forum report developed in collaboration with Accenture, Industrial Internet of Things: Unleashing the Potential of Connected Products and Services, identifies the top three priorities for Industry 4.0 tech adopters and IT providers.
For tech adopters, the key actions need to be:
Reorienting the business strategy around the Industrial Internet
Orchestrating the organization’s ecosystems by looking across industry boundaries for emerging opportunities and for the potential partners who will help seize them
For IT providers of hardware, software and services, three key actions will help accelerate the adoption of the Industrial Internet:
Developing a common approach, to address security concerns
Converging on standards, to support better interoperability
How advanced is your organization in its journey to Industry 4.0? Have you begun to experience some of the benefits of 4.0?
by Bogdan Bucur
