Amazon was pioneering the use of robots to order. Each day, FUSH carries on robots in fulfillment centers bellows full of heavy warehouses, and Asting employees with physically demanding tasks. These robots have come a long way since Amazon began inserting them over a decade ago, but there is still much to learn from how these system interacts with people.
Maya Cakmak, associate professor at the University of Washington (UW), helps explore these open questions. Through work supported via UW + Amazon Science Hub, a research collaboration located in UWS Allen School of Computer Science & Engineering is investigating Cakmak a generation of robots that can be trained not only to understand and handle heavy products but also for humans.
“I focus on making robots that can interact with people and perform everyday tasks,” she explains.
Cakmak helps tackle one of the most challenging parts of the fulfillment process: Picking and stewing. In a UW Robotics Lab that repeats a setup found in some Amazon fulfillment centers, CAKMAK and her team have built prototype robots that can extend robotic arms to compact storage areas, manipulate and grab objects and extract them from the mess with nearby objects.
The goal is to explore how robots and people can help someone else with challenging tasks while achieving the right result every time.
Human-robot interaction
Cakmak has been her career at the forefront-Human-Robot interaction or HRI-A field focusing on designing and building robots that can be “trained” and interact with people to help with specific, Somit’s personal, tasks. Her groundbreaking work started helping them with disabilities.
Maya cakmak discussion her research
“I focus most on robots performing physical tasks,” says Cakmak, “whether it is for the visual or hearing intaigned, those with cognitive challenges or physical impairments – everything that limits their access to the physical world.” She points to a popular root root swab as a good example, as it not only moves the environment but performs a task – vacuuming – it can be a task for anyone, but especially those with disabilities.
One of Cakmak’s most inspiring HRI projects was a robotic and software interface she designed for Henry Evans, who is Quadripiplegic and needs help performing everyday tasks in her home. With the help of just his eye movements and a few finger clicks, Henry can use an interface Cakmak helped design to instruct his home robot to help feed him and operate devices.
Over the last three summers, the interdisciplinary team contributes that Cakmak was part of Henry to feed himself independently, contribute to household tasks and even physically participate in social interactions, such as playing cards with his family or playing with his grandchildren. Working with Henry inspired new research projects in Cakmak’s laboratory, including the development of a tool set that enables the creation of fully customized interfaces that meet the specific needs and preferences of each unique user. The tool set was used in early stages of an Amazon project to quickly test the opportunity to tackle more challenging choices with a human operator in Løkken.
A passion for robotics
Cakmak, born in Belgium and grew up in Turkey, loved math from an early age competing in Math Olympiads while in high school. At the Middle East Technical University (METU), where she won a bachelor of electrical and electronic technique, she developed a passion for both technique and the idea of ​​using maths to solve problems in the real world. Her interest in robotics was inspired by a senior Capstone project involving the construction of a robot that climbed the stairs higher than themselves.
See the stepping climbing robot in action
This interest was elaborated as she continued her education at METU, completed a master’s degree in computer technology and led several robotic research projects. She moved to the United States in 2007 to pursue a Ph.D. in Robotik by Georgia Tech, specializing in Hri.
After completing his PhD, Cakmak was associated with robotists at Willow Garage, a groundbreaking robotics research laboratory and tech incubator that developed hardware and open source software for personal robotic applications. She originally joined as the “most senior trainee”, she quips, but the company quickly named her a post -doctoral research fellow, and she continued works on the real world robotics along with the sole of the best robotics in the field.
With a strong desire to remain in Akademia, Cakmak applied for professoring positions during his year at Willow Garage. She landed her dream job as assistant professor at Allen school in 2013 and was promoted to the lecturer in 2019, continued her work in HRI and helped inspire the next generation of roboticists.
Cakmak’s early research on UW focused on developing robots that could be taught or programmed by end users. While the procedure avoided the enormous challenge of building universally skilled robots, it also raised a more human-centered challenge: enabiling people who are not robotics or software developers to program robots. To achieve this, Cakmak and her team at Human-Centered Robotics Lab World close with end users to develop and evaluate new tools. The Laboratory’s Mission Stément calls for project that “focuses on robot programming of final use, use of robotic tools and Assive Robotics.”
Several years ago, the laboratory’s work began to draw notice from robotic scientists at Amazon. Cakmak’s expertise in HRI as well as her very different team’s entry into an Amazon Robotics Challenge, caught the attention of Michael Wolf, a head used scientist in Amazon Robotics. Cakmak’s work with human-to-robot software interfaces included the one she built for Henry made a big impression on the team.
“She is a great robotics in full stack, and she has a deep background in interaction between people and robot,” says Wolf. “These skills made her an excellent person to lead this research. We are engaged in the science community and do this in different ways. Supporting Maya’s work that has both science value and several potential social benefits is one of these”
Human-in-the-loop
Implementation of robots that can quickly identify a massive selection of items and greasy and remove individual items from tightly packed space is a huge challenge, especially on Amazon’s scale.
“It’s just inevitable that there will be corner cases and failures,” Cakmak said. “So how can we put people in the loop to reduce mistakes? It was my pitch that has a human component when we started the project.”
While Wolf notes that Prototype robots successfully can grease the majority of Amazon’s objects without errors, he and Cakmak Success in production will require greater cooperation between people and robots because of the pure diversity of ITMs in Amazon’s catalog.
Human-in-the-loop manipulation
Take books for examination. For efficiency, books are stacked closely together, making it challenging for robots to identify the right book, greased it and push it off the shelf. But with the help of a human telecommunications operator, the robot can gently slide out a book from above, making it easier to understand.
“There are restrictions on robots’ capabilities,” says Cakmak. “We have shown that human operators can use their finer perception and intelligence to help them perform more complex tasks.” The team is also working to develop a new method to enable robots to learn from human operators over time.
Broad impact
Wolf notes that the work that Cakmak and the Science Hub team perform has a lot of Broad applications, a fact that is in line with Hub’s mission to tackle harsh challenges in science and technique.
“When we start collaborating at Science Hub, we think that the technical holes are, but also where there is a super hard, real problem that may not get much attention,” observes Wolf. “Preferably reverberate these problem traits broadly with the academic society and excite the field beyond in collaboration.”
“We often talk about using robots for what they are good at and asking people to do what they are good at,” Wolf continues. “There is now this new cross that Cakmak helps define.”
In addition, Cakmak notes that the practical uses of the Science Hub Work Benefit Academics like her.
“I think that in robotics we have suffered from a culture to pose problems that can solve robots,” she says. “So you want to see robots playing Ping Ping Ping and using chopsticks. We like to challenge ourselves and do something cool without really thinking about applications. I really like that the Amazon work is group in real problem that will make a real difference.