ROS-Industrial for Real-World Solutions
| By: Tanya M. Anandan, Contributing Editor
Having mastered dull, repetitive jobs for decades, robots are ready for a challenge.
In the nation’s heartland, a smart robot decides in real-time how it will tackle a paint job where every single part is unique. Meanwhile, a robotic welder finds the seams on its own and welds an assembly it’s never seen before. Across the way in a warehouse, dozens of mobile robots nimbly fill orders with their own 3D printed hands, while in another part of the world, fleets of robots autonomously transport 1-ton loads of car parts across miles of factory floor without human intervention.
These aren’t farfetched robotic applications. They’re already here. And they all have one thing in common, ROS.
The Robot Operating System (ROS), developed and maintained by the nonprofit Open Robotics, is a collection of software libraries and tools that help developers and programmers build robot applications. From drivers and state-of-the-art algorithms to powerful developer tools, ROS is all open-source. Free to use, free to share.
The repository, which is a treasure chest of open-source software code and toolsets, is supported by Open Robotics and can be found at ros.org, with numerous additional and complementary repos found at GitHub. The hub is home to over 36 million developers working together to host and review code, manage projects and build software. GitHub is now owned by Microsoft Corporation.
ROS-Industrial (ROS-I) is an open-source project that extends the advanced capabilities of ROS to manufacturing automation. In addition to research and educational institutions that serve and collaborate with industry, ROS-I is particularly geared to robot manufacturers, component suppliers, system integrators and robot end users. No longer the sole domain of grad students huddled in university labs, or ambitious robotics incubators trying to change the world, ROS has become serious business in the industrial realm.
ROS provided a solid foundation. Now ROS-Industrial has come of age.
“As evidenced by the activity and sheer number of entrants to the order fulfillment, warehouse and logistics space, this area has proven that ROS-based solutions can survive and even thrive in manufacturing environments where uptime and reliability are critical,” says Matt Robinson, Program Manager for ROS-Industrial Consortium Americas.
High ROI, Focused Results
The ROS-Industrial Open-Source Project began as a collaborative endeavor between Southwest Research Institute (SwRI), robot manufacturer Yaskawa Motoman, and Willow Garage to support the use of ROS for robotics and manufacturing automation.
|First multi-purpose robot designed for use on the aerospace factory floor for aircraft sanding and coating removal using advanced sensors to conduct real-time path planning and analysis as it moves around the aircraft. (Courtesy of Southwest Research Institute)|
San Antonio, Texas-based SwRI is one of the oldest, independent nonprofit organizations in the U.S., providing innovative science, technology and engineering R&D services to government and industry around the world. Founded in 1947 by philanthropist and oilman Tom Slick Jr. on a converted cattle ranch, SwRI occupies 1,200 acres of laboratories, test facilities and offices for over 2,500 technical and support staff.
The Institute takes a multidisciplinary approach to problem solving, leading initiatives in fuel and energy efficiency, geosciences, turbomachinery, automated driving systems and energy storage. Over the years, SwRI has gained worldwide renown for leading several NASA missions.
SwRI, Yaskawa and Boeing recently collaborated on a ROS-Industrial project for the Advanced Automation for Agile Aerospace Applications (A5) Robotic System. Watch this 11-ton mobile robot demonstrate aircraft sanding and coating removal on a mock C-17 aircraft part.
ROS-Industrial was founded by former SwRI team member Shaun Edwards in January 2012. Managed by SwRI, the ROS-Industrial Consortium Americas launched in March 2013 and now has siblings in Asia and Europe.
ROS-I Consortium European Union is managed by Fraunhofer IPA in Stuttgart, Germany. ROS-I Asia Pacific is managed by the Advanced Remanufacturing and Technology Centre (ARTC) under the Agency for Science, Technology and Research (A*STAR), in collaboration with Nanyang Technological University in Singapore.
Together the three ROS-I Consortia, or RICs, are over 60 members strong, representing research and educational institutions, government, startups and large multinationals around the globe. Members come from nearly every major industry, including automotive, aerospace, construction, agriculture, logistics, electronics, power generation, healthcare, defense, process automation and robotics. Big or small, these are significant industry players.
RIC members are forward-thinking, innovative companies populated with thought leaders and visionaries who are investing in their organizations’ futures as well as industry’s future as a whole. Return on investment is high. Each member pays for only a fraction of the development efforts, but each benefits from all of the results.
Members work collaboratively to develop an application roadmap for ROS-Industrial, set near-term technical goals, and participate in spinoff Focused Technical Projects (FTPs) to develop tools and application-specific software capabilities of interest to multiple members.
Windows and ROS 2.0
Sticking close to its open-source roots, ROS was originally developed on the Linux operating system. But now Microsoft has joined the open-source initiative to broaden the exposure of ROS and ROS-Industrial to an even larger community of users and contributors at home in the Windows ecosystem, and to further the digitalization of industry through cloud, Internet of Things (IoT) and AI solutions.
Building on the success and widespread adoption of ROS, Open Robotics officially released ROS 2 in 2017. The next-generation development framework will include support for multi-robot systems, real-time control, Windows and other platforms, enhanced security and standardization, and more use cases in real-world production environments.
Microsoft support and the ROS 2 release will only enhance the opportunities ROS-Industrial affords robot OEMs, integrators and users. As the ROS to ROS 2 transition progresses, the ROS-Industrial community will continue to provide guidance as new versions are released.
Within ROS and the surrounding open-source ecosystem are packages that enable intelligent application development, including:
~ 2D/3D point cloud processing
~ Robot motion planning and navigation
~ Offline visualization and planning tools
We’ll see these tools put to task as ROS-I Consortium members share use cases for high-mix robotic painting, mobile robots for factory intralogistics, and robot-to-goods picking solutions.
First, let’s address some misconceptions about ROS, and explain what ROS can and cannot do.
Myth #1 - ROS Not Safe to Download/Use
Because it’s the source code, ROS is significantly more secure than any canned software you would download on your PC, according to SwRI’s Robinson. You can see every character, every call and every keystroke. There’s no hidden file you can’t read in open-source.
In fact, ROS-Industrial focuses on code quality and reliability because of their importance to the manufacturing automation world.
“It’s well known in the IT world that the big players like IBM download open-source code and then wrap their IP around it,” says Robinson. “Nobody starts from scratch. That’s the idea we are trying to bring to industrial. Of course, nothing replaces proper validation. Good runoffs, good system testing, software hardening.”
Myth #2 – Open-Source Means No IP
One of the first misconceptions people have when they hear “open-source” is that they will lose all of their IP. Not the case.
Consortium members say ROS is very open (even for open-source, due to user-friendly licensing), in the sense that you can take and use software packages and develop modules and code, do what you want with it, build on top of it, and then repackage it in your own proprietary way. You don’t have to resubmit it to anyone, and you don’t have to divulge it to anyone if you choose not to share your experiences with the community.
ROS gives a lot and asks for little, if anything, in return. And it’s the gift that keeps on giving.
Myth #3 – Open-Source is Unreliable
Some might assume open-source means immature, unstable and unreliable. According to Consortium members, that’s not necessarily true. You may occasionally run across a piece of code that might be buggy, but it’s not prevalent. In fact, they find it to be the opposite.
“It’s a bit like crowdsourcing,” says Spirit AeroSystems’ Curtis Richardson. “There are lots of people who love to spend time trying to stress-test software code and they’ll find the bugs for you, even when you thought your own code was perfect. You get free support and improvement opportunities that you might not have the time or resources to do yourself.”
ROS is essentially a middleware or a framework that enables efficient communication between disparate pieces of hardware and software libraries. ROS-Industrial takes ROS and extends it to manufacturing-relevant hardware or industrial-relevant use cases and applications.
“If I develop a system with a robot, a camera and other components, it may work great,” says Robinson. “But if I try to change out the robot, I’m essentially starting from scratch. The advantage of starting with a ROS-based approach is that I have a middleware that abstracts away all the details of the specific hardware I choose.”
He likens it to a Windows driver on your computer desktop. The first time you plug in a new mouse, Windows has to download a driver so the device works properly.
ROS removes the potential constraints posed by your hardware selection, or helps you avoid having to reinvent the wheel should you have to go from one robot brand to another brand, or from a particular sensor to a different brand or type of sensor.
It’s important to note that just because a company “uses” ROS, doesn’t mean that it’s in the final product. There’s a difference between using ROS as a foundation for development versus actually running a solution with ROS as the middleware.
“Once you get to a development standpoint where you don’t need hardware interoperability and/or distributed components, you can back out the ROS components and make a traditional hardened software stack,” says Robinson. “Many companies have done this.”
Fewer Barriers with ROS Drivers
ROS has been around for over 10 years. There’s an extensive community of users who are always contributing significant improvements to ROS and the complementary tools. Many OEMs are now offering ROS drivers or interfaces so their hardware plays nice with the ROS ecosystem right out of the box.
|Robotic 3D printing on a non-planar surface using open-source tools for generating robot trajectories with a ROS-enabled robot. (Courtesy of University of Southern California)|
Yaskawa Motoman was one of the first industrial robot OEMs to develop ROS drivers for its robots. These drivers allow the robot controller to communicate with ROS. All Yaskawa robots are ROS-enabled. It should be no surprise that many pioneering ROS-based technologies feature Yaskawa robots.
“Not everyone can get access to an industrial robot,” says Robinson. “It’s difficult to optimize a driver or do application development without access to an industrial robot. Yaskawa has broken down a lot of barriers to let people get their hands on Yaskawa robots.”
Robot manufacturer ABB Robotics is another ROS-I Consortium member supporting ROS drivers for its robots. Many 3D sensor manufacturers, such as Consortium member Pilz, also provide drivers to the ROS community.
There are ROS drivers for many of the robots on the market, but they are developed by the community. As such, every time a robot OEM updates a controller, the community-developed drivers may no longer work as anticipated.
“We would like the robot OEMs to recognize the value of participating in the community and develop and support their own drivers,” says Robinson.
He says other robot manufacturers are starting to come around. “They see these students in the universities doing sophisticated research and using their robots for advanced capabilities, realizing these students will end up at the Googles, Amazons, Microsofts and innovative manufacturing software companies of the world.”
As those students, who are now accustomed to the interoperability that ROS affords them, emerge from the universities, they will put more pressure on the OEMs. Today’s students will become the decision-makers of tomorrow.
ROS-I the Accelerator
Yaskawa was on board with ROS right from the start, helping launch the ROS-Industrial Open-Source Project.
“We realized there was a real need to make robots smarter, and to do so, it required a community, not a village,” says Roger Christian, Divisional Leader, New Business Development at Yaskawa Motoman in Miamisburg, Ohio. “The fact that SwRI decided to shepherd ROS-Industrial added a lot of credibility to this open-source concept. I understand Microsoft is now collaborating with ROS, which will continue to strengthen its appeal and make using ROS more attractive to our integration partners.”
Christian works with universities to support the advancement of robotic capabilities through Yaskawa’s membership in the ARM Institute. Watch this 3D printing application developed with open-source tools and a ROS-enabled Yaskawa robot at the University of Southern California.
“To his credit, Erik Nieves convinced Yaskawa to get involved in the ROS-I community,” says Christian. “He visited universities from Boston to Silicon Valley and discovered they were all using ROS for advanced robotic development, so Yaskawa developed a robust ROS communication driver, which allows our robots to be ROS-compatible and development-friendly. That’s how it all started.”
Nieves, the former technology director at Yaskawa, is now CEO and Cofounder of startup Plus One Robotics. He realized early on the value of ROS for accelerating and enabling industrial robotics applications development. Many of the talking points in Nieves’ ROSCon 2013 presentation are still relevant today.
Christian says ROS is an accelerator. “ROS is the preferred development software environment for any university that offers advanced robotics degrees. You can bet the grad students leaving to create the next revolution in robotics start their development with ROS tools. It offers a large and growing set of programming tools that can accelerate their desired end-product development without requiring months of backend core-coding.”
|Leveraging open-source tools, proprietary AI and sensors, and ROS-enabled hardware, this startup’s proposed solution for high-mix, low-volume robotic welding claims to locate part seams and generate robot motions on the fly without costly, labor-intensive programming. (Courtesy of Path Robotics)|
He also works with startups that initiated development in ROS-I to create easy-to-use solutions for welding, material handling and logistics. Many of these emerging technologies have the potential to lower barriers for small and midsized companies that may want to embrace robotics but don’t typically have the resources for costly, labor-intensive process development and programming.
“Yaskawa is committed to offering the best interface for anyone wanting to develop advanced robotics technology in ROS,” says Christian. “Our team is making sure our ROS driver remains full-featured, robust and accessible to all developers.”
Diverse Industry Expertise, Similar Challenges
Spirit AeroSystems was also an early adopter of ROS-Industrial. The aerostructure manufacturer and supplier for major OEMs such as Boeing, Airbus and Bell Helicopter, has made significant contributions to the ROS-I community.
“Over the last few years in particular, there’s been a groundswell of credibility for ROS-Industrial,” says Curtis Richardson, Technical Fellow for Spirit AeroSystems in Wichita, Kansas. “The more people that join, the more momentum it gets, and the more credibility it has. Now it’s considered cutting-edge.”
Richardson says the ROS-I Consortium is a truly collaborative environment.
“We get to sit at the table with colleagues who are developers and experts, and have similar interests and objectives. One of the greatest things is that it’s not just aerospace. It’s wonderful to hear that folks from Caterpillar and BMW, and all of these other industry segments, share similar challenges. They have a lot of similar perspectives on what they would like to see and the capabilities they need. We can collaboratively work towards addressing those challenges.
“We know we can learn a lot from the automotive industry, but equally from the heavy industries, the Ag guys, etc. We all help each other.”
High-Mix/Low-Volume, ROS-I to the Rescue
ROS-Industrial has several software suites of feature-rich tools that lend themselves to highly dynamic environments where real-time adaptability is imperative. According to users in the Consortium, there are few if any alternatives on the commercial market to handle these types of high-mix, low-volume applications.
“Aerospace, in a general sense, tends toward the high-mix, low-volume scenario, especially compared to industries like automotive,” explains Richardson. “That traditionally creates a type of barrier to entry for automation. ROS is one of many tools available to help attack that.
“You combine things like machine vision, AI, cobots and mobile robots, and ROS is one of those tools that helps bring all those things together in a cost-effective way and helps shorten development timelines. Applying those types of technologies and capabilities to low-volume, high-mix challenges has been our primary focus.”
|Open-source software tools for real-time robot trajectory planning use 3D scan data to identify the condition of a part and create a robot program on the fly. (Courtesy of Spirit AeroSystems)|
Spirit AeroSystems’ first foray into deploying ROS-Industrial was nearly five years ago on a high-mix robotic painting application for commercial aircraft skins. Admittedly, they were not deep experts in ROS development, so the aerospace supplier partnered with SwRI to develop technology for identifying the condition of a part and then creating a robot program on the fly. The result was one of the first implementations of what would become Scan-N-Plan™ technologies, a suite of ROS-I software tools that enable real-time robot trajectory planning from 3D scan data.
The Scan-N-Plan approach is particularly effective for these types of conditions:
~ Highly variable part mixes where hand programming is impractical
~ No CAD part models available
~ Flexible or deformable parts, making pre-programming impossible
~ Part-to-part variability that is difficult to accommodate with static programming
~ Applications requiring flexible part fixturing or no fixturing at all
Using robotic automation to paint these large commercial aircraft skins is particularly challenging. The aluminum sheeting is relatively shiny, so Richardson says it’s difficult to apply sensor technology to it. Using sensor fusion, the ROS-I tools combine all the data from different sensors and create the process solution for that particular part.
Even more challenging, every single part is different.
“The ROS-I piece is essentially telling the robots what part they have and what condition the part is in, so that they know where to go and what to do for that particular part,” explains Richardson. “Every part that comes through has a brand-new, never-before-used robot program applied to it. We don’t build a program, store it and then the next time that part comes through, recognize it and pull up the right program. No, we actually create a brand-new program every time a part comes into the paint booth.”
This painting process has been done manually for decades. For that reason, the parts never had to be consistently loaded into a tool. No one ever had to worry about what shape the part was in.
“The aircraft skin is pretty thin, so it can deform very easily,” says Richardson. “When it’s on the picture frame tool that it rides into the paint booth, it can be loaded differently each and every time. It can be pulled out of contour very easily. The ROS-I tool in effect reverse-engineers that part and tells the robot what it looks like in reality.”
|Two inverted robots paint commercial aircraft skins using sensor fusion and open-source software tools for real-time robot trajectory planning on the fly. (Courtesy of Spirit AeroSystems)|
“We developed and proved the capability together with SwRI,” says Richardson. “Then Spirit AeroSystems took the responsibility to finalize the code and package it, and then implement it and maintain it.”
Painting was the first commercial application of Scan-N-Plan, but with improvements over time, the ROS-I technology has become applicable to other processes, including surface finishing, deburring, inspection and surface blending.
Spirit AeroSystems has other ROS-based systems in the works, and just kicked off another project using ROS-I in collaboration with SwRI, Wichita State University, and funded by the ARM Institute that will use collaborative robots for sanding aerospace parts.
Richardson says ROS-Industrial is part of their long-term strategy. Watch innovation take flight at Spirit AeroSystems.
“We really want to take the Scan-N-Plan work we’ve done and move it closer to state of the art in machine learning and artificial intelligence, and using perception with mobility and navigation,” he says. “We’re also working on human interfacing, like integrating augmented and mixed reality. Those things together we think will hold a lot of potential for our safety strategy in terms of getting automation more deeply embedded in our manufacturing.”
From the aerospace industry to the automotive market, robot users are leveraging ROS-Industrial to help them lower costs, achieve greater efficiencies, and reach coveted milestones in their Industry 4.0 strategies.
Between its four brands – BMW, BMW Motorad, MINI and Rolls-Royce – the BMW Group produced more than 2.5 million vehicles in 2017, an all-time high for the German automaker. Every day, suppliers deliver over 31 million parts to the 30 BMW production sites worldwide. To ensure the company’s global production network receives parts in a timely and reliable manner, BMW continues to expand its use of innovative technologies in production logistics.
In 2015, BMW began developing a self-driving Smart Transport Robot (STR) for transporting roll containers through production halls. The low-profile mobile robots autonomously calculate the ideal route and move freely through the factory space, carrying roll containers weighing up to one ton. A built-in battery module from the BMW i3 all-electric sedan powers the STR for a full work shift.
|Autonomous mobile robot uses sensor fusion capabilities developed with ROS open-source software tools to freely and safely navigate the production aisleways of this automaker. (Courtesy of BMW Group)|
ROS-I tools were used as a foundation for STR development. BMW also partnered with Microsoft on the cloud-based fleet management hub for the autonomous transport systems. Last year, SwRI’s Matt Robinson gave Microsoft a tour of the Institute’s robotics lab in exchange for an onsite preview into ROS-I development applications at BMW Group.
Check out this globe-spanning episode of Microsoft’s “DECODED show” and see the Smart Transport Robots on the job at BMW.
With ROS-I, BMW Group Logistics was able to incorporate several different sensors into their STR to enable sensor fusion within each mobile robot. ROS-I allows the different hardware and sensors contained within the vehicle to communicate with each other, and also allows the robot to communicate with different IoT solutions. BMW uses a cloud-based operating platform for central coordination of the STRs.
This interoperability is a key benefit for many large robot users. It’s the main reason BMW chose ROS-I over commercial solutions.
Carlos Ferretiz is the Logistics Expert in Robotics at BMW Group in Munich, Germany. He’s focused on ROS-I and the automaker’s Smart Transport Robots.
“ROS enables us to introduce all these IoT SDK packages from different companies to communicate in our cloud system,” explains Ferretiz. “The cloud system is in charge of sending the job to the robot. To introduce this system into different platforms from other companies would have been quite difficult, because it’s a closed system and we are not allowed to go into the system and import these packages. That’s why we decided to start our platform with ROS.
“It’s easy to introduce different packages and test different solutions, and then decide which solution is the best,” he continues. “Using ROS with different platforms, different robots and different sensors is really fast,” noting this speeds overall development of the platform.
Rich Toolsets, Community Support
The ROS-I community’s support and collaborative software development model was also a driving factor in BMW’s STR development process. They use ROS-I toolsets for navigation, visualization and simulation.
“We started using TurtleBots (open-source mobile robot research platform) to help us understand how we can use the navigation stack from ROS and how we can implement it in our industry,” says Ferretiz. “One of the benefits is the support we get from the ROS community. We learn from their experiences and how they have been implementing different packages. This helps us to adapt them to our solution.”
He says the different forums within the community provide a platform for discussions between different ROS developers to not only help solve specific development problems, but also inspire new capabilities development.
“All of these rich toolsets for development and simulation that ROS offers helps us understand all these new technologies,” says Ferretiz. “The tool that helped us the most is Rviz, which is a 3D visualization tool. We use it specifically for our navigation stack to visualize all driving behaviors.”
|Autonomous mobile robot receives its orders from the cloud and transports roll containers across the plant, making sure the automaker’s production lines receive parts reliably and on time. (Courtesy of BMW Group)|
Rviz also logs sensor data from the robot, which helps BMW understand why some “missions” don’t go as expected.
“We call them missions when the robot is going from one point to another, or when the robot is transporting some material from one point to another,” says Ferretiz. “When there is a failure in a mission, in order to log information, we use what is called Rosbag, where we save all the information from the laser sensors and the odometry from the motors. We can see this information in Rviz and then we have a complete understanding of why the robot is misbehaving in an area. We can react quickly to solve these kinds of errors.”
Martin Bauer is Product Owner of Services for Autonomous Transportation Systems (ATS) at BMW. He leads a team of almost 30 people developing the platform BMW will use to connect and integrate all autonomous vehicles.
“The openness of ROS and connecting it to our new open platform for ATS helped a lot because we learned how to standardize that interface as well, and define it,” says Bauer.
Open-source simulation tools allowed BMW to test fleet management capabilities before deploying large fleets of robots in its plants. Simulation will also act as a virtual proving ground for future deployments.
“We used ARGoS to program the simulation of our service solution for more than 100 robots,” says Bauer. “We started with Gazebo, but when we scaled up, we had to move to ARGoS.” He says ARGoS scales better than Gazebo, especially for large numbers of robots.
BMW has five plants using STRs in production areas. Before production deployment, each robot goes through pilot testing and then staging, before joining the general factory population.
“In the following years, the robot fleet will increase significantly so that we have over 3,000 autonomous robots in our plants,” says Ferretiz.
IoT and Cloud Solutions
Like many ROS users, BMW is looking forward to Microsoft integration. BMW already had a close partnership with the software giant, as the plants have over 3,000 machines running on the Microsoft Azure industrial IoT cloud platform.
“At the moment, our ROS tools are running on Linux,” says Bauer. “But we are working with Microsoft to find out how far Windows is on ROS in regards to our solution and what additional benefits Windows can bring.
“Our cloud solution runs on Microsoft Azure, and we have been working very closely together with various Microsoft development teams, for example building a robot simulation which is available as an open source component,” he says. “Now we are formalizing that partnership on the Open Manufacturing Platform.” (OMP is a new initiative to drive open industrial IoT development and help foster a community to build future Industry 4.0 solutions.)
BMW envisions production facilities where the best suitable robot for a specific transportation task can be freely chosen by the plant. Hence, they need an open platform that gives them the capabilities to design and execute various use cases by providing generic and lightweight components.
ROS-Industrial joins a long legacy of robotics technology that continues to transform the automotive industry. But perhaps no other sector has benefited more than the logistics space, where developers use ROS-Industrial to deliver innovative solutions for their customers. ROS helps keep e-commerce on pace with demand.
The intralogistics market is expanding at a frenetic pace. Mobile robots, picking robots and other types of material handling and logistics automation are in high demand by the always-on supply chain.
Bastian Solutions, a material handling and robotics system integrator headquartered in Indianapolis, Indiana, provides automated solutions for distribution, manufacturing and order fulfillment centers around the world. Bastian is a ROS-I power user, and Chief Innovation Officer Chris Morgan is a fervent advocate.
|Developed using ROS open-source tools, robotic mobile manipulator picks multiple product types and batches with greater efficiency. (Courtesy of Bastian Solutions)|
“ROS is such an incredible framework,” says Morgan. “When you’re doing these things at high volume for customers, you need a highly dependable framework that you can go to at a moment’s notice and leverage those pieces of technology and bring them together very rapidly. If we have to write lines of code for every little thing out there to achieve these goals, we wouldn’t be nimble and agile, and be able to help the customer to get where they need to go quickly.”
Bastian Solutions was acquired by Toyota Industries Corporation (TICO) in 2017 and is now part of Toyota Advanced Logistics North America, created by TICO to complement its material handling side of the business. In his role at Bastian Solutions, Morgan brings emerging technologies to Toyota and its partners.
“We have many pieces of technology that Toyota, Vanderlande, Raymond and Bastian Solutions make across the globe. My team not only forecasts the future, we also guide where those investments should go for the next-level technologies that will help solve customer problems. Along the way, we help develop immediate-need technologies. We also partner very closely with all of the other business units across the Toyota ecosystem to help them come up with the latest technologies.”
Scorpion is one of those technologies. This supersized mobile manipulator moves robotic batch picking to the next level. Integrating a mobile platform with a vision-enabled material handling robot, Scorpion serves as a “robot-to-goods” solution by picking multiple product types, eaches and SKUs for various orders at a time. See it in action.
Developed using ROS-I as a foundation, the Scorpion is in its pilot stages. Bastian expects it to be production-ready later this year.
Silo Buster, Solid Foundation
Morgan says ROS-I is a silo buster. ROS-I breaks down silos internally within companies and externally across different organizations and industries.
“We’re trying to figure out how to glue different pieces of technology together,” he says. “Internally in a company, different groups will choose different ways of doing things. ROS helps pull them together. It helps us accelerate our development, build a better robot for that unique brand. Then it takes that approach and unifies it for all of the developers within the community.
“It helps us all speak the same language, regardless of what part of the world we’re in, so we can help enable each other to build better products across the globe.”
|Autonomous robotic shuttle is equipped with a 3D printed arm and gripper destined for e-commerce robotic picking applications. (Courtesy of Bastian Solutions)|
Bastian Solutions’ new autonomous robotic shuttle was developed in the ROS vernacular. The shuttle uses a robotic arm and custom gripper with 3D printed components made possible through a collaboration with additive manufacturing solutions company Fast Radius. Take a look behind the scenes.
The lightweight but durable polymer arm will allow the agile robotic picker to quickly navigate a grid of totes, picking product as it goes, for an entirely new paradigm in robotic picking solutions.
“You don’t have to bring the tote to a person to pick the goods. The entire order is picked within the grid and then it’s brought out for delivery,” says Morgan. “The shuttle can also leave the grid entirely and take products anywhere in the warehouse. It can also repack and replenish the grid by filling up totes and then the shuttle drives back into the grid.”
The patented system made its debut at ProMat 2019. When released, the shuttle will target the e-commerce market and retailers struggling to keep up with the Amazon effect. That includes grocery stores now offering one-hour curbside pickup.
“We’re trying to save all the brick-and-mortar stores,” says Morgan. “I want the Meijers, the Walmarts, the Targets of the world to survive.”
Morgan says they used open-source tools like Gazebo, Rviz and MoveIt! in the shuttle’s development process. “Then we used our own mixed blend of ROS, ROS-I and proprietary code base to make this all come to life. But you start with ROS as a foundation. Whether we’re developing this new shuttle, the Scorpion mobile manipulator, our new ULTRA trailer loader, or a Raymond truck, we use a variety of tools for vision, navigation and mobility – those are all in the ROS toolbox.”
A wise man once said, “ROS itself is not the value. The applications it enables are the principal value.” (Thank you, Erik Nieves)
We can’t wait to see what ROS-Industrial enables next!