The National Science Foundation (NSF) awarded more than $2 million to fund projects led by Georgia Tech robotics researchers. The principal investigators (PIs) and co-PIs for these projects represent three of the Institute’s six colleges, illustrating the interdisciplinary collaboration that distinguishes Tech as a leader in the national initiative to accelerate the development and use of robots in the United States.
Georgia Tech faculty have a strong tradition of exceptional research and a robust interdisciplinary focus . Extremely proud of and continually impressed with the contributions our researchers make to advancing robotics
Three projects received NSF funding through the National Robotics Initiative program, which was unveiled by President Obama in June 2011, and is led by NSF with support from NASA, the National Institutes of Health, and the United States Department of Agriculture. Tech’s new projects focus on the development of the next generation of robotics and the advancement of the capability and usability of such systems in innovative application areas:
- “Learning from Demonstration for Cloud Robotics”—Led by School of Interactive Computing Associate Professor Andrea Thomaz, this project received $426K and aims to leverage cloud computing to enable robots to efficiently learn from remote human domain experts.
- “Understanding Neuromuscular Adaptations in Human-Robot Physical Interaction for Adaptive Robot Coworkers”—Led by School of Mechanical Engineering Assistant Professor Jun Ueda, this research focuses on developing theories, methods, and tools to understand the mechanisms of neuromotor adaptation in human-robot physical interaction. Associate Professor Minoru Shinohara (School of Applied Physiology) and Assistant Professor Karen Feigh (School of Aerospace Engineering) serve as co-PIs on the project, which received almost $1.2M.
- “Don’t Read My Face: Tackling the Challenges of Facial Masking in Parkinson’s Disease Rehabilitation through Co-Robot Mediators”—Led by College of Computing Associate Dean & Regents’ Professor, Ronald Arkin, this project received almost $580K and has two primary goals: 1) developing a robotic architecture endowed with moral emotional control mechanisms, abstract moral reasoning, and theory of mind sensitive to human affect and ethics; and 2) creating a specific architecture for a robot to mediate communication barriers between caregivers and patients with Parkinson’s disease who experience “facial masking,” or lack of recognizable emotion.
The fourth project, “Bioinspired Collaborative Sensing with Novel Gliding Robotic Fish,” received more than $83K from the NSF’s Robust Intelligence (RI) program, which encompasses all aspects of the computational understanding and modeling of intelligence in complex, realistic contexts. Led by School of Electrical & Computer Engineering Associate Professor Fumin Zhang, the research aims to establish a theoretical framework and provide an enabling technology for robust underwater collaborative sensing with small, inexpensive robots.
Robotics research at Tech attracts more than $35 million in sponsored research each year. Core research areas include mechanisms, control, perception, artificial intelligence, human interaction, and application technologies. The Institute continues to advance personal and everyday robotics through its research into the ways robots can learn from and interact with humans, and by exploring issues surrounding their governance and ethical use.
Great story in NY Times today about the cooperation between humans and robots. The next generation of robot system systems will be Co-robots where they cooperate with humans to perform tasks that may be difficult to perform by humans alone as they require heavy lifting, high precision or the tasks are highly repetitive. Humans are still incredible in terms of perception, dexterity, cognition and reasoning, so the combined human-robot system offers a number of added advantages.
One of the challenges is also is also to provide easy programming. A traditional industrial system will have a cost break down of ~25% for the robot, ~25% for auxiliary hardware and 50% for software. Through new programming paradigms it is possible to design systems that are much faster to program – the Baxter robot is an example of a new generation of such systems.
As mentioned in an earlier posting the US National Robotics Roadmap was published the past week. The roadmap is a revision of the First US Robotics Roadmap that was released May 2009 based on a CCC sponsored study. The second version of the roadmap contained an update to three sections i) manufacturing, ii) healthcare/medical robotics, and iii) service applications (domestic and professional). In addition, new sections covering defense and space were added to the roadmap.
During 2011 we saw a 40+% increased in robot sales in the US for manufacturing. We also saw significant growth to service and healthcare applications. Overall the sector experienced fantastic growth. We have also seen how utilization of robotics and automation has enabled companies such as Apple, Lenovo, GE, Foxconn, … to setup new manufacturing facilities on US soil. Robotics has become an important catalyst to drive forward jobs, the economy and building stronger communities. An important challenge is to ensure education of our workforce. This includes all levels of the enterprise from design to manufacturing and from factory floor to board room. We have a significantly shortage of people to staff the manufacturing enterprise.
Press release from Georgia Tech – March 20:
Robots are being used more widely than expected in a variety of sectors, and the trend is likely to continue with robotics becoming as ubiquitous as computer technology over the next 15 years.
That is the message Henrik Christensen, Georgia Tech’s KUKA Chair of Robotics in the College of Computing, will bring to the Congressional Robotics Caucus on March 20 as he presents, “A Roadmap for U.S. Robotics: From Internet to Robotics – 2013 Edition.”
The report, which outlines the progress of robots in multiple industries over the last five years and identifies goals for the coming decade, highlights robotics as a key economic enabler with the potential to transform U.S. society.
“Robots have the potential to bring manufacturing jobs back to the U.S., to improve our quality of life and to make sure our first responders and warfighters stay safe,” said Christensen, who is also the coordinator of Robotics VO, sponsor of the report. “We need to address the technical and educational needs so we can continue to be leaders in developing and using robotic technology.”
A group of more than 160 experts from universities, industry and government came together for five workshops over the last year to fully evaluate the use of robotics across various applications and create a roadmap to the future. Christensen is presenting that report to lawmakers as a guide on how to allocate resources to maximize progress.
Most notably, the group found using robots in manufacturing could help generate production systems that are economically competitive to outsourcing to countries with lower wages.
Already companies like Apple, Lenovo, Samsung and Foxconn have begun to “reshore” manufacturing by using robotics in production systems. The sale of robotics in manufacturing grew by 44 percent in 2011 as robots have become cheaper and safer. The use of robots is shifting from big companies like GM, Ford, Boeing and Lockheed Martin to small and medium-sized enterprises to enable burst manufacturing for one-off products, the report found.
But Christensen notes that automation in manufacturing will not lead to job losses for U.S. workers, but will create new high-value jobs.
“Some jobs will be eliminated but they are the ‘dirty, dull and dangerous’ jobs,” Christensen said. “Those jobs will be replaced with skilled labor positions. That’s why one of the goals in the roadmap is to educate the workforce.”
In addition to manufacturing, robots are helping businesses, such as Amazon, improve logistics and reduce delivery costs, a savings that could be passed onto the consumer. In agriculture, robots are being used to precisely deliver pesticide onto crops, reducing unnecessary exposure of chemicals on produce. The report recommends that progress in both areas be expanded.
With advances in human-like manipulation, robots are increasingly assisting individuals with disabilities with tasks like getting out and preparing meals. They are also being used in 40 percent more medical procedures than a few years ago in a greater number of surgical areas such as cardiothoracic, gynecology, urology, orthopedics, and neurology. The use of robots for surgery can reduce the complications by 80 percent, the report found.
Robots have proven their value in removing first-responders and soldiers from immediate danger. More than 25,000 robotic systems were deployed in Iraq and Afghanistan for ground and aerial missions. More than 50 percent of pilots in the U.S. Air Force operate remotely piloted systems and never leave the ground.
Robots are also becoming integral part of space exploration, such as Opportunity and Curiosity on Mars. A “robonaut” is on the International Space Station helping with menial but important research tasks.
As impressive as the progress in robotics has been, the report outlines 5-, 10- and 15-year goals to take robotics to the next level. Critical capabilities that should be developed for robotics include 3-D perception, intuitive human-robot interaction and safe robot behavior.
The report is an update to the initial robotics roadmap, which was published and presented to Congress in May 2009. That roadmap led to the creation of the National Robotics Initiative, an effort jointly sponsored by the National Science Foundation, the U.S. Department of Agriculture, the National Aeronautics and Space Administration and the National Institutes of Health. It also established Robotics VO, an umbrella organization that brings all robotics players together to focus on joint initiatives.
“Robotics is one of a few technologies capable of building new companies, creating new jobs and addressing a number of issues of national importance,” said Christensen. “We hope this report will help foster the discussion on how we can build partnerships and allocate resources to move the robotics industry forward.”
Recently there has been a lot of press stories about robots, the economy
and its relation to employment. The stories are mixed. Some claim that
we will loose jobs to robots others are saying that through use of
robots manufacturing in the US will grow. In reality both of those
viewpoints are valid and understandable. Unfortunately most of the
stories are presented as one or the other. A more two-sides presentation
Over the last few years we have seen an increase in use of robot
systems. Robots have been used for dull, dirty and dangerous jobs. This
includes welding, sorting in warehouses, electronics assembly. It is
clear that introduction of robots to perform a job that a human was
doing before might displace a person. There are three kinds of uses: 1.
introduction of robots to perform tasks that cannot be performed by
humans, 2. introduction of systems to assist people, and 3. introduction
of systems to do work formerly performed by humans. The motivation for
displacement of labor is typically higher performance, which can be
lower price, higher quality or improved performance.
Throughout history there has been a worry about new technology. About
130 years ago people were worried with the introduction of steam engines
and assembly lines. What would happen to the labor force? Eventually we
had the automotive industry that could generate affordable cars for
everyone and a significant number of new jobs. Around 1980 secretaries
were worried about their jobs as typing pools disappeared due to the
introduction of personal computers. Today most people handle their own
email, … but there are nonetheless more administrative assistants
doing travel, scheduling, meeting minutes. The jobs have become
increasing skilled. So one of the challenges we are experiencing is that
some unskilled jobs gradually are getting automated and the displacement
in jobs is from unskilled to skilled labor. It will be important for the
work force to continue to receive training.
As an example through use of robots it is possible to reduce the price
of manufacturing, which in turn enables in-sourcing. Apple has
announced a new plant in the US and so has Lenovo. Tesla manufactures
green cars in California. Welding is performed by robots just as it has
been done for 30 years in Detroit. Some of the jobs carried out in
foreign countries can only be insourced through use of automation.
However, every job in manufacturing creates another 1.3 associated jobs
in services, supply chain, etc. The economy is growing and slowly we are
recovering jobs. The current trend is that jobs that were outsourced
earlier are returning to America, some of them through use of automation
and some new ones for manufacturing and some in associated industries.
The annoucement was broadcast on the NASA public feed. The game is termed ultimate ascent. The field will have two pyramids and the robots will be required to use friesbees to score on the buckets at the end of the fields. Double points for scoring during the autonomous period. At the end the robots can climb the pyramids to score more points. The code is sAucersFlyRobotsClimb!
We had 53 teams at the launch event at Georgia Tech. The event was organized by the RoboJackets and sponsored by the Woodroof School of Mechanical Engineering, RIM@GT, GM, National Instruments, United Technologies, College of Computign, Kimberly Clark, Automation Direct, MSC Industrial Supply. More than 800 high school students showed up for the event.
Good luck to everyone for the 2013 season and remember to have fun.
A new year is here. Welcome to 2013.
The time around New Year is a great time to reflect on how far we came in 2012 and where we might be going in 2013. Looking at this from a professional point of view I think robotics is a very exciting place to be.
2012 we had a number of great things happen:
- The joint NSF, USDA, NIH, and NASA National Robotics Initiative saw its first set of awards. In total more than 700 proposal were received, which were reviewed in 20+ panels. The request for funding was close to $1B and with an projected budget of $45m it was no surprise that success rates on proposals was low. However, we now have officially a set of ~30 projects that are funded under the NRI. I am sure we will see many proposals submitted for the 2012/2013 round.
- A new organization the US Robotics Virtual Organization or robotics-vo for short was launched. This is a national robotics network similar in spirit to the European Robotics Network – EURON, that was launched around 2000. The network is trying to coordinate – a roadmap for robotics in the US, educational resources, best practise for technlogy transfer, and a press club for dissemination of information about robotics.
- As one of the first efforts the Robotics-VO has setup a set of five workshops on roadmapping. This in turn has enabled an update of the US National Robotics Roadmap. The old roadmap from 2009 was in need up an update and many things has happened since then. In addition there was a need to augment the roadmap with consider military/first responder needs and also to align the roadmap with the NASA agenda. All of this has been accomplished and shortly (February) the revised roadmap will be published. A briefing to the congressional caucus on robotics has also been planned.
- During 2012 we saw some major commercial successes. KIVA was sold to Amazon for 700m+ which clearly illustrates the potential for use of robot technology in logistics. An area we can expect to see further growth in during 2013.
- The year 2012 we also saw the public announcement of the first robot Baxter from Rethink Robotics (former Heartland Robotics). A two armed robot that is considered safe for use in human environments at a price of less than $25k is a major achievement. It appears to be well suited for simple pick and place operations. It will be interesting to get a hands-on experience to see how well it does in real applications. With a higher speed it could be very interesting for logistics applications. The stiffness could be a challenge for real assembly operations, but it will be interesting to test it. Also a developer API is supposed to surface shortly for academic users.
- There are by now a fair number of dual arm manipulator systems and given a mobile platform it is only natural DARPA launched the Disaster Robotics Challenge, where teams use humanoid platforms to demonstrate performance for first responder type scenarios. Given what we saw at Fukoshima in Japan during March 2011 this is a very natural and timely opportunity.
- Apple announced that they will start manufacturing the next iMac line of computers in the US. The fact that we have started to in-source is a big deal. Through use of automation we can close the barrier between manufacturing with low salary workers and smart manufacturing systems. Others such as Tesla have decided from the outset that manufacturing will be in the US.
- Willow Garage spun-out their perception work in Industrial Perception and the ROS effort was made independent in Open Source Robotics Foundation, and other systems such as the Point Cloud LIbrary and OpenCV was also made into independent entities. An industrial version of ROS was also launched through the South Western Research Institute.
For 2013 there is no doubt that we will see a number of new interesting opportunities
- The National Robotics Iniative will continue to grow and as more agencies become active players in the program there is no doubt we can build sustainability, growth and longer-term perspectives. It will be important to see further engagement of industry to make sure that new R&D efforts lead to results that are commercialized. The objective is clearly to try to at least have a budget of $100m for 2013/2014.
- The first Robotics-VO PI meeting will take place and it will be a great opportunity to get a broader sense of what is contained in the program and also to try to engage industry in transition of results into real products / processes
- The first results from the DARPA DRC will be shown. Initially it will be in simulation, which will be a good start.
- More and more companies such as Motoman, Rethink Robotics, Schunk, Yujin, … are providing two armed manipulation system. It will be exciting to see new applications with these systems in manufacturing, logistics and service applications. The real challenge is now in the integration of these systems into applications
- For the application of robots it will be interesting to monitor the Industrial ROS effort. Traditionally industry has had a hard time embracing open source. There are a number of challenges in terms of stable releases, a unified architecture, proper code reviews, etc that must be adopted to make these systems reliable enough to be used in major manufacturing systems. However this challenge has been overcome before. Excellent examples include Linux, GNU (sub-systems), … Through consideration of best practise in these areas there is no doubt that robot systems integration can arrive at a similar place, which could lead to a new degree of economic growth due to lower price of deployment and a higher degree of interoperability.
- The EU is launching a new framework program by the end of the year. The new program is named Horizon 2020. The most relevant program is in the cognitive systems and robotics division. The program enable broader international collaboration (INCO) and the initial focus will be around inclusion of US partners in new projects. That is – US universities and companies – can participate as equal partners in the projects and also be paid by the EU as part of a projects. In this past this has been possible in theory but in reality it has been a major challenge to make this happen.
These are merely a few of the things we can look forward to in 2013. This is going to be another exciting year! Happy New Year to Everyone.
The website for the US Robotics Virtual Organization - Robotics-VO is now live. The website has a number of useful tools. It covers the progress on the update of the US National Roadmap for Robotics. It has a fairly comprehensive calendar with calls for proposals, conferences, deadlines for papers, and we are starting to see educational material emerge aswell. Please check out the site. Much of the information is only available after you register. For now only people in the US can register to use the site (sorry).
Recently there has been a lot of discussion in the media about driverless cars or more appropriately autonomous cars. That is, cars where the driver does not have to have his/her hands on the steering wheel. A number of competitions have taken place to evaluate the performance of current technology. Early work both in the USA and in Europe resulted in cars that drove in traffic over extended distances. The DARPA Grand challenge saw 5 teams complete a desert course of more than 50 miles by 2005. Two years later the DARPA urban challenge saw 6 teams complete a 60 miles city like course with traffic . The DARPA Urban Grand Challenge motivated companies such as Google to pursue design of cars that can handle traffic, respect the traffic law to the letter and interact with other entities such as pedestrians, bikes, etc. on the road. Today legislation has been passed in the state of Nevada to allow use of driverless cars as part of regular traffic. In addition, both Florida and California have initiated efforts to revise their traffic laws to allow use of driverless cars. There is no doubt that there is a need for serious political considerations to make this possible across all areas as also noted by Brian Albright – “Driverless Cars – a politically hot potato”
Other news stories have indicated that driverless cars only can be deployed in combination with significant investments in infrastructure. The argument is that such an investment would be too significant and either reduce deployment or prevent deployment entirely. The use of common infrastructure such as embedded landmarks/beacons is not a new idea. Such an approach was tested as part of the California Partners for Advanced Transportation TecHnology (PATH) project
about 15 years. The approach is technically feasible, as it was used for convoys of cars to travel down 20 miles of highway, but the solution is economically unrealistic. However, we have seen tremendous progress on new sensor technology, computing and robust algorithms, which in reality alleviates the need for such fixed infrastructure.
We have seen steady progress on embedded new technology into cars. Already today some cars will warn you if there is another car in your blind spot. Some cars have technology to monitor if you brake the best possible way to avoid a collision. These are all ways of slowly augmenting the cars to improve safety and relieve drivers of some of the aspects of driving. Such a gradual introduction of technology is likely to be important to long-term adoption of the technology. There is no doubt driverless cars is the way of the future. It allows us to reduce the number of accidents, we can utilize infrastructure more effectively, and drivers can use commute time to do texting, read email, and whatever else would be a more productive use of their time. In the US alone more than 30,000 people are killed in traffic accidents every year. Our infra-structure has at best an 10% utilization, which could be doubled through use of driverless cars. According to US Department of Transportation the average American spends close to one hour per day commuting to / from work. Using such time productively could have a significant potential for the individual and society.
The introduction of driverless cars will allow people to continue to use a cars as they age, it will be safer, more economical, we will have better productivity, etc. All in all the driverless cars is a great investment for society and for individuals. However, it must be recognized that this will be an incremental process so it will take time before such solutions are deployed across multiple nations and across multiple brands of cars. The progress see with the Google Driverless cars from the Google X-Labs and the recently reported drive from Parma, Italy to Shanghai, China are both great examples of how far technology has progress and an clear indication of things to come.
About a year ago IFR (International Federation of Robotics) contracted the company Metra Martech to study the impact of robotics on employment. Typically people predict that introduction of robots result in loss of jobs. Recent publications such as the “Make it in America” by Andrew Liveris, CEO and Chairman of Dow Chemical have suggested that through adoption of advanced manufacturing technologies the industrialized nations can compete with countries where low-salary workers are responsible for the manufacturing.
The recent Metra Martech study estimates that close to 3 million jobs today are enabled by use of 1 million robots. In addition the report predicts that over the next five years another 1 million jobs will be created due to adoption of robotics technology for applications in consumer electronics, solar & wind, and advanced fuel cell technology.
Both Japan and Germany are leaders in use of robotics technology and this has resulted in increased employment in sectors such as automotive, that traditionally have been heavy users of robotics technology.
The report predicts that robots will continue to be major players in automation of factories, but that the new application areas will include elderly care and medical applications. In addition homeland security and defense will maintain its position as a high value market.