A big reason for this is the work that researchers are doing in the computer science department at the University of North Carolina. Drs. Ron Alterovitz, Dinesh Manocha and Ming Lin are a few of the brilliant computer scientists in Chapel Hill working on algorithms that will change the way we interact with robots, making them smarter and our lives better. Their research in planning algorithms will one day make robots capable of doing housework, working with people on factory floors, assisting in surgery and driving our cars.
Through their work with GAMMA, Manocha and Lin are developing motion-planning algorithms that will make it possible for robots to navigate complex environments. Companies like German manufacturing giant KUKA work with GAMMA to figure out how best to implement these algorithms in their robots.
Places like homes and factories are changing and complex environments. While we already have robots in these environments performing repetitive tasks like vaccuming (the Roomba was introduced in 2002), the next generation of robots must learn, respond and adapt to changes in the environment and interact with humans.
The challenge is making the robots capable of learning on their own and anticipating human behavior (and without realizing some fear that they could overtake humans). People react to one another without thinking about it—it’s our innate instinct developed over millions of years of evolution. Robotics is a relatively new field and it will take some time for scientists to re-create in machines what nature has in humans.
Take for example, the home, where robots will one day help the elderly and people with disabilities.
“The goal here is to be able to have an assistive robot— a robot that can move around in the environment but also can manipulate objects,” says Alterovitz, who leads the Computational Robotics Research Group at UNC, which develops algorithms for assistive robots among other projects.
While the research is a long way from producing a Rosie, a robot that can interact with and learn from humans the way other humans do, in the near future we will have robots capable of accomplishing simpler tasks around the home – think picking up dropped utensils for an elderly person.
The main challenge is developing algorithms that will make robots capable of learning new tasks in unfamiliar environments, much like humans are capable of doing.
“Instead of having to program every little detail in the robot, we’ll program in the ability of the robot to learn and then somebody like an occupational therapist can demonstrate to the robot how to do certain tasks. The robot should be able to autonomously perform those tasks even if the environment is a little bit different,” says Alterovitz, who has received awards from the National Science Foundation and the National Institutes of Health for his work. Here's a video of his early work on this concept using a robot made by the Boston startup, Rethink Robotics:
Through the algorithms Alterovitz and his team are developing, robots will be able to learn to perform new tasks from human beings. These algorithms, combined with collision avoidance technology, will allow the robot to act autonomously and perform tasks even when new obstacles are added to the environment.
So, can the robots do our laundry?
“Something like that,” Alterovitz says with a smile. “It’s going to be a while before that works in people's homes.”
Right now, the computational robotics research team at UNC is building up from smaller tasks such as wiping tables and scooping sugar. While this may seem unimpressive at first glance, the difference a robot with these capabilities could make in the life of a disabled or elderly person is significant.
But Alterovitz says there’s still a big gap between where the laboratory research is and having robots performing these tasks in people's homes.
So where are robots already making a difference? Here's a sampling of UNC's work (and that of other industry leaders) in a variety of fields.
Robot Life in the Factory
Robots are already on our factory floors. Large and mid-sized manufactures are already using machines to perform repetitive tasks on assembly lines. Dr. Manocha imagines a future that is a bit different.
“Robots have been used in manufacturing; that’s not new,” says Manocha, another NSF award winner. “A big trend is coming where robots are going to become like co-robots. We are heading to a stage where robots and humans could be working together.”
While he’s hesitant to say when robot coworkers will be commonplace, Manocha and his colleagues are confident it will happen one day. Indeed, companies like Rethink Robotics (founded by the Roomba creator) have already made great strides in making safe, affordable and efficient robots for factory floors.
As a sidenote, water cooler conversations are going to be less lively, but on the plus side, you won't need to worry about your robot coworkers cornering you with pictures of their children.
The impact this can have on the economy is enormous. The proliferation of robots in factories has the potential to bring manufacturing jobs back to the states by reducing production costs, increasing efficiency and creating demand for a more skilled and educated workforce. Studies already show that robotics help create jobs, not replace them.
“Robots are already making a difference to the share of labour in total costs. Cheaper, more user-friendly and more dexterous robots are currently spreading into factories around the world, and they cost just the same in America as they do in China.”
Robots in Surgery
Dr. Alterovitz and the Computational Robotics Research Group are also working with physicians and engineers at UNC and Vanderbilt University to develop algorithms that will allow robots to assist surgeons during certain operations. With funding from the National Science Foundation and the National Institutes of Health, they are developing steerable needles and concentric tube robots that can move around anatomical obstacles in the human body to deliver drugs and perform surgeries less invasively than possible with straight needles.
“We’re developing planning-and-control algorithms for the robot to be able to do that sort of steering and this allows the physician to focus on the higher level aspects of the procedure,” says Alterovitz.
This technology will improve patient outcomes in certain types of surgeries, such as pituitary gland operations. Traditionally these surgeries have been extremely invasive, forcing surgeons to damage healthy tissue to reach a tumor and often making it difficult to remove the entire mass. Concentric tube robots will allow surgeons to take a less invasive route, lowering the risk for the patient and allowing physicians to operate in places of the body that are difficult or impossible to reach.
Robots on the Road
If you’re not sufficiently freaked out by this point of the article, let’s talk about driverless cars.
While we may be a long way off from having flying cars like George Jetson’s available at the local Hyundai dealership (we can already imagine the irritating Sport Durst commercials), driverless cars may soon be ubiquitous on our nation’s streets and highways. This is a bit concerning when you consider Siri sometimes has trouble directing me to the closest Cook-Out to my house.
“It’s coming whether we like it or not,” says Lin, who helped develop some of the first algorithms for real-time collision detection in the 1990s at the University of California at Berkeley. They've since been used in computer game development and computer-aided design. Her research at UNC has contributed to the advancement of technology that allows driverless cars and robots to avoid obstacles and adjust course.
Driverless cars, or as Lin calls them “autonomous ground vehicles” will one day be driving us to work and delivering cargo for shipping companies. In the future, human drivers will share the road with robots. While the technology exists, it will be a few years before we see driverless cars on the road. Google X is hoping to have driverless cars on the road by 2020—one of its key researchers is former UNC post-doctoral student and GAMMA researcher Jur van den Berg.
As with assistive robots and manufacturing robots, there’s still some work to be done before robots will be able to fully anticipate human behavior.
“As we design robots we need to be able to also better model human behaviors because robots are going to be everywhere and they are going to be interacting with humans,” says Lin. “The robots need to be able to learn how to anticipate human behaviors.”
The same algorithms that Ming and her colleagues are developing to help robots navigate roads can also be used to help human beings navigate and better inform our decisions regarding civil infrastructure such as highways, roads, bridges, public transit and airports.
According to Lin, 90% of our nation’s civil infrastructure is underutilized. She says we can combine the traffic information we are already collecting with algorithms designed for robots to make more efficient transportation.
The Proliferation of Robots
As the science evolves, robots are going to become a bigger part of our day-to-day lives. They will drive us to work, help Grandma out around the house and assist surgeons in the operating room. The rise of the machines has the potential to change our lives for the better.