After turning the dial to heavy-duty and pushing the start button, Modi waits for the click that indicates that the interlock mechanism successfully locked the door of the washing machine. Shortly after, the water starts to flow and the machine indicates that the soak and scour phase is in progress. Now it’s time for Modi to move on to the next washing machine.
Modi is short for MoDiBot, or Mobile Diagnosis Robot, an autonomous test platform for household appliance testing, developed by engineers at Loccioni, a leading provider of test, measurement, and automation solutions in Italy. MoDiBot works in a test laboratory, allowing scientists and researchers to apply new technologies and validate the algorithms and applications of their autonomous systems to put them on the touchstone for adoption in the lifecycle test laboratories of white goods manufacturers around the world.
Without a doubt, robotics technology is an emerging priority. While in the past, robots were mainly used for repetitive tasks in industrial manufacturing, today they play an important role in applications characterized with one of the 3 Ds: dull, dirty, or dangerous. Thus, it is no surprise that robotic technology is adopted for applications in homeland security, inspection, exploration, healthcare, logistics, and even the entertainment industry.
Dr. Cristina Cristalli, Director of the Research for Innovation Group at Loccioni, focuses on robots for mobile testing and their flexibility and multifunctionality. According to Dr. Cristalli, robotics is an excellent way to move and place sensors and measurement equipment, opening a new perspective for the automation of reliability testing. While there are some challenges, most of the robotics systems used across industries and application areas share the same core components and architectural approach. The combination of one or more robotic manipulators (arms) mounted on top of an autonomous platform is a common mechanical solution. Based on this combination, the industrial and research communities have developed many application-specific variations over the last few decades; however, robots are far from being engrained in our everyday life, and robotics technology has not been adopted at the same rate as information technologies or RF technology. So, what are the technologies that pave the way for robots out of the laboratory?
A significant part of the innovation takes place in the “smarts”of the robot, where a common architecture includes a network of interconnected embedded systems. Some of these embedded systems perform small and dedicated tasks such as battery management, motor control, or sensor fusion, and are based on lower-footprint programmable logic devices (PLDs) and microprocessors. Additionally, most robotic systems contain one or more high-performance embedded systems that use the latest processor technology to execute mission- critical tasks, ensure safety, perform image processing, and run advanced control algorithms. Overall, robots are complex mechatronic systems, and programming them often requires a large team of PhD-level experts.
Another question is what technologies are driving robotics innovation and will transform robotics into a standard engineering discipline, allowing a broader adoption. The big innovations are in the embedded control systems, software tools, sensor technology, and battery and energy efficiency technology. No matter how complex the hardware architecture might be, Sense, Think, and Act is a common paradigm to implement innovative robotic applications.
A big impact on the adoption of robotics comes through the ever increasing processor performance. Only recently available in desktop PCs, powerful multicore-processors have been adapted for embedded systems. These processors are attractive for robotics systems because lower-footprint embedded systems comply with the tight power, space, and weight constraints of the robotics industry. Powered by real-time operating systems, they are capable of executing many of the mission-critical tasks, close control loops, and can perform I/O operations and communicate with other devices. Additionally, the robotics industry is adapting PLDs such as Field Programmable Gate Arrays (FPGAs) for tasks that require custom hardware, extreme speed and reliability, or parallel execution. These off-the-shelf systems come equipped with a set of middleware and drivers to abstract the complexity of the hardware, and do not require embedded design expertise. Other systems, like the National Instruments NI CompactRIO, allow customers to add I/O, communication, motion, and vision capabilities based on their application requirements and start working on their software development right away.
The computational components alone, however, will not simplify the development of robotics applications. According to Dr. David Barrett, roboticist and professor at Olin College, “The robotics industry badly needs an industrial- grade, hardened, richly supported software development system to build intelligent, autonomous, mobile robots that can sense, think, and act in the complex real world around them.” In other words, industry needs a software tool that abstracts the complexity of these interconnected high-performance computational devices and provides ready-touse building blocks for the most common tasks in robotics applications such as sensor connectivity, navigation, localization, path planning, obstacle avoidance, or vision-guided motion.
Over the last couple of years, industry and the research community developed these types of tools in the form of opensource software packages like Willow Garages Robot Operating System (ROS) or commercial tools like NI LabVIEW Robotics, a graphical system design tool that supports different models of computation for the implementation of robotics applications and makes it possible to deploy to different targets such as multicore processors, real-time processors, or FPGAs. With these tools, engineers are ready to adopt robotics technology for the next generation of their products, similar to the Loccioni engineers, who were able to come up with a revolutionary concept for the lifecycle test of white goods.
Sensors and Batteries
Besides embedded control systems and high-productivity software tools, sensors and sensor connectivity play a key role; thus, it is not surprising that the current proliferation of sensors is shaking up the robotics industry. The availability of compact MEMS (microelectromechanical system) sensors that are heavily used in commercial products such as smart phones or controllers for entertainment consoles have led to a price erosion for accel - erometers, gyro sensors, and even more standard sensors including light, pressure, or temperature. But it is not solely about price. The MEMS technology also simplifies the processing of sensor information because the devices include signal conditioning and translate the sensor information into useful data that is sent back via standard protocols such as I2C.
The reuse of commercial technologies sometimes goes even further. Rather than capitalizing on price advantages for components used in commercial products, the robotics industry adapts the commercial product directly. USB cameras or the popular Kinect controller from Microsoft are good examples that demonstrate this trend. Recently, this commercial product made it possible for many robotic systems to recognize and locate objects in a 3D space at a significantly lower cost than previous solutions.
The last key area of innovation to deploy robots to the field is the area of energy efficiency and battery technology. Operated in a research laboratory, robotic systems usually stay connected to a power source. While this might work for some application areas, there are many use cases of robotics where the autonomous operation requires independent energy sources, often in the form of batteries. Implementing energy efficient operation of the robot actuators and using low-power embedded control systems can help reduce energy consumption. At the same time, there has been a lot of innovation in battery technology, which is driven by the focus of the automotive industry on alternative energy sources for modern vehicles and the increasing use of mobile devices like laptops, tablets, and smart phones. The robotics industry can capitalize on the heavy investment of global players such as Toyota, GM, Apple, or Samsung.
All of these technologies were essential components in the design and deployment of MoDiBot to the first lifecycle test facility. Robotic technology allowed the manufacturer to perform automated testing of dozens of devices at the same time and around the clock. The same technologies that enabled Loccioni to develop their revolutionary test system will help to grow the adoption of smarter robots in industrial and commercial applications, and will transform robotics technology into a commodity engineering discipline rather than simply a research exercise. Ultimately, this will make it possible for robots to leave the laboratory and settle into our everyday lives.
This article was written by Christian Fritz, Senior Product Manager for Advanced Machine Control and Robotics at National Instruments in Austin, TX. For more information, Click Here