The wonder that is IKEA has brought us many delights over the years – beautifully decorated showrooms, those tasty Swedish meatballs, and mostly easy to build furniture that, if not itself 3D printed, can at least be augmented with the technology. But not everyone thinks the brand’s furniture is simple to assemble, which is why they’ll be happy to hear that there’s now a robot to take care of it.
After three years of research, which were supported by grants from Singapore’s Ministry of Education, the Singapore-MIT Alliance for Research & Technology (SMART), and NTUitive, the innovation and enterprise arm of Nanyang Technology University (NTU), a team of researchers from the university’s School of Mechanical and Aerospace Engineering have developed an autonomous, furniture-assembling robot that’s made up of two robotic arms with grippers and a 3D camera.
Last month, the NTU team, comprised of Dr. Francisco Suárez-Ruiz, university alumnus Xian Zhou, and assistant professor Quang-Cuong Pham, published a paper in the Science Robotics journal on the results of their robot, titled “Can robots assemble an IKEA chair?“
“For a robot, putting together an IKEA chair with such precision is more complex than it looks. The job of assembly, which may come naturally to humans, has to be broken down into different steps, such as identifying where the different chair parts are, the force required to grip the parts, and making sure the robotic arms move without colliding into each other. Through considerable engineering effort, we developed algorithms that will enable the robot to take the necessary steps to assemble the chair on its own,” explained Professor Pham.
“We are looking to integrate more artificial intelligence into this approach to make the robot more autonomous so it can learn the different steps of assembling a chair through human demonstration or by reading the instruction manual, or even from an image of the assembled product.”
In just 8 minutes and 55 seconds, the robot, made with components purchased off the shelf, was able to assemble one of IKEA’s classic Stefan chairs, made of lacquered pine; this included finding all of the parts and independently planning its motion pathways. So how did the NTU team manage to build a robot that can accomplish a task so odious that some couples even consider divorce afterwards?
The robot was specifically designed to mimic the “hardware” of people used to put things together – the arms are industrial, six-axis motion robotic ones, with two parallel grippers for picking up objects and two wrists with force sensors, so the robot knows how strongly it’s gripping something and pushing it into contact with another object. The eyes are a detailed 3D camera, which have been used in the past to help bring next-level capabilities to the virtual world of gaming, capture historically significant buildings and artifacts, and even help in surgery.First, the robot takes 3D photos of the parts spread out, in order to quickly work up a reliable map of their estimated positions on the ground. According to NTU, this helps replicate “the cluttered environment after humans unbox and prepare to put together a build-it-yourself chair.”
Coding algorithms from three open source libraries, designed by the team, help the robot plan a quick, collision-free motion with both “hands,” which is also integrated with both tactile and visual perception. The amount of force used by the robotic arms to grip pieces and perform simple tasks, like inserting those little wooden plugs, has to be regulated. As Professor Pham explained, this is difficult to do because industrial robots are better at precise positioning than they are force regulation.
The force sensors on the robot’s wrists help the robot consistently, and accurately, pick up the pieces, as well as sliding the plugs along the surface to find the holes and then inserting them.
While the thought of a robot building my furniture is certainly appealing, its greater purpose is far more important. It’s being used to study and explore an area of robotics known as dexterous manipulation, which needs precise force control and resembles humans in its object manipulation, as it actually has special robotic fingers and hands.
Professor Pham said that up until this point, autonomous demonstration of dexterous manipulation has only been used on more “elementary tasks.”
“One reason could be that complex manipulation tasks in human environments require many different skills,” Professor Pham explained. “This includes being able to map the exact locations of the items, plan a collision-free motion path, and control the amount of force required. On top of these skills, you have to be able to manage their complex interactions between the robot and the environment.
“The way we have built our robot, from the parallel grippers to the force sensors on the wrists, all work towards manipulating objects in a way humans would.”
The NTU researchers believe that their autonomous robot could, as the university put it, “be of greatest value” when used to perform precise tasks in industries that don’t require specialized machines or assembly lines, though robotics is making its mark there as well.
Now that they’ve reached their goal of demonstrating the robot’s ability to build an IKEA chair, the team is partnering up with various companies to apply this type of manipulation to multiple industries. For example, they’re currently working to prepare the robot for glass bonding – helpful in automotive applications – and drilling holes in metal aircraft components.
Discuss this and other 3D printing topics at 3DPrintBoard.com or share your thoughts in the comments below.
You May Also Like
4D Printing in China: Shape Memory Polymers and Continuous Carbon Fiber
Researchers have been looking further into the benefits of shape memory polymers (SMPs) with the addition of raw materials in the form of continuous carbon fiber (CCF). Authors Xinxin Shen,...
3D Printed Wireless Biosystems for Monitoring Cerebral Aneurysms in Real Time
Continuing to further the progress between 3D printing and electronics within the medical field, authors Robert Herbert, Saswat Mishra, Hyo-Ryoung Lim, Hyoungsuk Yoo, and Woon-Hong Yeo explore a new method...
Feasibility Models to Determine Efficacy of 3D Printing Over Traditional Methods
In ‘Model for Evaluating Additive Manufacturing Feasibility in End-Use Production,’ authors Matt Ahtiluoto, Asko Uolevi Ellman, and Eric Coatenea encourage the idea of exploring 3D printing for designs first, comparing...
Refining Macro and Microscopic Topology Optimization for AM Processes
Researchers from Italy and Germany continue along the path so many are following in refining and perfecting 3D printing processes. In the recently published ‘Structural multiscale topology optimization with stress...
View our broad assortment of in house and third party products.