ROBOCUP@HOME SOCIAL ROBOT TEAM
Social Robots are a disruptive innovation poised to change lives, and transform business and society. The University of Technology Sydney (UTS) is Australia's research leader in Social Robotics.
UTS researchers, graduate and undergraduate students work with industry and government partners on the implications and applications of social robotics. UTS organised the 2014 International Conference on Social Robotics in Sydney, serve on the Editorial Board of the International Journal for Social Robotics and program committees of the leading conferences in Social Robotics, AI and Robotics, and Human-Robot Interaction.
UTS Unleashed! Team designs and develops innovative software on Pepper Robots.
UTS Unleashed! is the only Australian Team that qualified for the 2017 RoboCup@Home Social Robotics competition, which will be held in Nagoya, Japan.
Our focus is innovative social behavours for robots that enhance human interaction and engagement. We are leading a workshop on human-robot engagement at IJCAI-17 in Melbourne with our industry partners CBA and Softbank Robotics and a Special Session at the RO-MAN 2017 on Intention in Social Robotics with CBA, Leeds University and Carnegie Mellon University.
Why UTS created a new RoboCup Team
UTS is engaging in the RoboCup@Home Standard Social Platform competition for several key reasons:
- to engage with a global strategic scientific effort that seeks to make safe and intelligent social robots a reality.
- to make major scientific contributions to the rapidly advancing field of social robotics,
- to build National capability in Social Robotics, an emerging disruptive technology, that will impact every Australian industry, and
- to provide transformational leadership opportunities and transdisciplinary practice-based learning experiences for UTS students.
UTS students will have the opportunity to:
- lead the design and development of an innovative transdisciplinary social robot software system in collaboration with an industry partner, Softbank Robotics.
- engage in discovery and transdisciplinary research and innovation.
- experiment, test, evaluate and reflect on their transdisciplinary solutions using benchmark challenges at an international competition.
- benchmark their work with other students from world leading international universities.
- contribute to an aspirational international movement designed to shift the social robotics frontier and advance the field of robotics to change peoples lives and transform business and society.
- develop new and transformational capabilities and skills in: designing an entirely new system for a law-abiding social robot that must interact with people safely and effectively while facing challenges in a home scenario requires sophisticated transdisciplinary creativity. Some examples include the following:
- design thinking and innovation: designing an entirely new system for a law-abiding social robot that must interact with people safely and effectively while facing challenges in a home scenario requires sophisticated transdisciplinary creativity.
- predictive real-time data analytics: robots are essentially big data analytics systems: using predictive data analytics to build models for real-time decisions making with copious volumes of complex unstructured data in multiple forms collected through their distributed sensors.
- transdisciplinary design and complex problem solving: designing social robot behaviours and application requires a rich transdisciplinary mindset. Researchers and students from different disciplines including Law, Health, Business, Psychology, Cognitive Science, Computer Science and Engineering, work jointly to create a new understanding of social robots in business and society.
- collaboration: we work as a team focused on the challenging practical problem of designing and developing an integrated social robot so students learn to be accountable and adaptable, how to build trust, negotiate, and persuade.
- leadership and entrepreneurship: every team member is expected to lead; to be strategic and entrepreneurial; to inspire other team members to develop high levels of confidence enabling high-orbit innovation and impact.
- Professor Mary-Anne Williams [Team Leader]: Mary-Anne is Director of the Magic Lab, Australia's leading social robotics lab. Mary-Anne has been Team Leader of more than 10 internationally competitive RoboCup teams since 2002. She serves on the Editorial Board for the International Journal of Social Robots; program committee for International Conference on Social Robotics and International Conference on Human-Robot Interaction; was Conference Chair, International Conference on Social Robotics 2014; served as Judge for RoboCup@Home Open Challenge, and leading the organisation of an IJCAI-17 Workshop on Human-Robot Engagement and 2017 RO-MAN Special Session on Intention in Social Robotics.
- Jesse Clark: Senior researcher and software engineer who spent 7 years at NASA leading the development of highly sophisticated systems including data management systems for the International Space Station.
- Dr Benjamin Johnston: University Medalist, Chancellor's Postdoctoral Fellow and winner of the 2010 Ray Kurzweil Prize for the most creative idea in AI. Key member and software development lead for two RoboCup teams. Program Chair of the International Conference on Social Robotics 2014.
- Dr Xun Wang: IBM Fellow, Chancellor's Postdoctoral Fellow, Senior Robotics Engineer at the Commonwealth Bank. Key member and software development led for two RoboCup teams.
- Dr Richard Billingsley: Senior Research Associate with expertise in Deep Learning and Robotics.
- Meg Tonkin, PhD Student focusing on social robots and the Internet of Things. Meg recently won the 2016 Best Robot Design Prize at the International Social Robotics Competition.
- Sammy Pfeiffer: Senior software engineer and PhD student who developed software on the humanoid robots at PAL Robotics. Team leader of PAL Robotics Robocup@Home in 2014 and key member of the two previous Robocup@Home editions. Also conducted several @Home demonstrations at RoboCup competitions.
- Jonathan Vitale: PhD student focusing on human social cognition, with a particular interest in developing models inspired by embodied cognition theories for robots. His research interests cover: face perception and processing; attention and executive functioning; models of emotional interactions between humans and robots; Theory of Mind.
- Le Kang complete his First Class Honours degree in computer science in the Magic Lab 2016 gaining a Social Robotics intership with Stockland. He commenced his PhD in 2017 to focus on Human-Robot Engagement focusing on HRI via Pepper's display.
- Suman Ojha: PhD Student focusing on robot emotions and ethical decision making.
- Sidra Alam: PhD Student focusing on robots as colleagues in the workplace.
- Siva Leela Krishna Chand Gudi: PhD student focussing on Social Robotics. Before joining UTS, he worked for various research organizations including the Indian Space Research Organization(ISRO), Defence Research and Development Organization(DRDO), Indian Railways and Bharat Sanchar Nigam Limited (BSNL).
- Syed Ali Raza: PhD Student working on learning by imitation. Ali is a founding member of the Karachi Koalas and has lead the development of two RoboCup 3D Simulation Teams.
- Thomas MacKenzie: Bachelor of Business student working on the distruptive innovation aspects of social robotics.
- Sophie Phillips: Bachelor of Creative Intelligence and Bachelor fo Business student working on the distruptive innovation aspects of social robotics.
- Several other UTS students will join the team along the journey to RoboCup 2017 and beyond....
Scientific approaches and achievements include:
- Social Human-Robot Interaction: Robot pointing; Robot-human collaboration; Human-robot affordance; Motion and gesture learning; Robot learning by human demonstration.
- Social Robots from Research Lab and into the Wild: We work closely with our industry partners. Our social robot systems have been released in the wild at the Commonwealth Bank Innovation Lab in Sydney and several Stockland Shopping Malls. You can download the mobile app, NSW Government Innov8, that let's you see Gutsy dancing in virtual reality, more than 3000 people engaged with Gusty in vitual reality at the Sydney Royal Easter Show in 2016.
- Novel Algorithms for Social Automonous Mobile Robots: Navigation with glass walls; PR2 Inverse Kinematics Engine; people tracker for mobile social robots.
- Novel Intelligent Social Robot Maneuvers: Robot soccer dodge; robot soccer dribble; robot soccer passing; robot soccer collaborative localisation using the ball; robot soccer strategy using case based reasoning.
- Pioneering approaches and implementations of Belief Revision for Intelligent Robots: VADER, SATAN and ADS.
- Innovative Attention based Cognitive Architecture for Social Robots: ASMO, the ABC Architecture.
- Innovative Social Robot Programming Paradigms: PyRIDE, SIMKIT, JoyRIDE
- Innovative Commonsense Reasoning Framework: COMIRIT.
Sample videos that highlight the skills of UTS Unleashed!
UTS has developed novel and innovative systems for a range of social robots. UTS is home to Australia's only Personal Robot 2 (PR2), GUTSY, one of the world's leading social robots. We are key members of the PR2 Research Community, which includes Stanford University, University of California at Berkeley, MIT, Brown, Carnegie Mellon University, Cornell University, University of Texas, Georgia Tech, University of Washington, University of British Columbia, University of Tokyo, Chinese Academy of Sciences, KU Leuven, Samsung, BMCE Bank, and Bosch. In addition, we work closely with industry partners in Australia, e.g. the Commonwealth Bank of Australia. We also advise government on innovation and entrepreneurship.
This social robotics video received the Most Entertaining Autonomous Robotics Video Award at the International Joint Conference on AI 2013.
Social robots dancing with humans. It turns out that dancing is a sophisticated human cognitive capability just like human language.
GUTSY using our our world first open source reverse kinematic engine to copy any motion. In this case, handwriting that demonstrates screen-to-hand coordination.
GUTSY is the first PR2 robot to be able to understand human language using IBM Watson technology.
We developed the UTS Unleashed! PyRIDE system to advance social robot rapid prototyping capabilities to help students accelerating learning.
UTS Social Robot System running on the Pepper robot during an invited visit to the University of Tokyo, December 2015.
GUTSY, Australia's only PR2 Robot, making coffee - a proof of concept, answering the question "Can Gutsy make coffee".
GUTSY is the first PR2 robot to be able to use Google Home technology to perform voice commands.
GUTSY using virtual reality! Another world first!
Moving around a modern building and developing social protocols for interacting with people in and around elevators is one of our research use cases.
UTS Unleashed! developed one of the most powerful kicks in the RoboCup SPL soccer league. The Ninja Kick can be used to score from anywhere on the field.
UTS Unleashed! developed the first Dodge in the Legged League to develop a dodge. The video contains the first dodge ever executed by UTS in a match against CArnegie Mellon (CMU). The CMU goalie makes a great clearing kick and a great save.
Human to PR2 Robot Motion Imitation and Recall System. Developed by 4th Year UTS student Michelle Yousseff.
A public performance of the UTS Song and Dance. To dance is human, to sing and dance is divine.
The Karach Koalas are a highly successful RoboCup 3D simulation team and a joint venture between UTS and IBA. This video was prepared just before RoboCup 2014. It shows clips from games with the University of Texas at Austin, University of Paris and Oxford University.
On the left is a video of the first ever robot dodge in the legged league. UTS used this manouver in a game with CMU at RoboCup 2004.
Previous participation in RoboCup
UTS has led numerous highly innovative and competitive RoboCup teams to championships since 2003 across four RoboCup leagues: Special Platform Soccer (AIBO and NAO robot leagues), 3D Soccer Simulation, Robot Rescue, and RoboCup@Home.
The team name UTS Unleashed! was original inspired by our first robot, the puppy-like Sony AIBO robot that we used to play soccer at RoboCup 2003. UTS Unleashed! evolved to encapsulate our vision for intelligent, autonomous, proactive, social, and untethered robots designed to enrich people's lives.
1. RoboCup Soccer
Team: UTS Unleashed! competed at RoboCup 2003, 2004 and 2010 using AIBO and NAO Humanoid robots. This team was the top team in 2004 and the Australian Robot Soccer Champions. It holds the record for scoring the most goals in a game 16-0 at RoboCup 2004.
Team: WrightEagle Unleashed! was established in 2008 as a result of a joint venture between UTS and the prestigious University of Science and Technology of China (USTC). It was the first team that not only spanned two countries but two continents. USTC developed China's first computer and is the only University founded and managed by the Chinese Academy of Science. It has been consistently ranked as the top University in China by Nature Index NPI 2012 - 2014. The team defied all odds and were Runners-Up in 2008.
Team:NuBOTS Mary-Anne established this team in 2001 at the University of Newcastle while in the Faculty of Business and Law in close collaboration with colleagues and students in the Faculty of Engineering. She was the Team Leader at RoboCup 2002. This team took 3rd place in both the Soccer Matches and the Scientific Challenges, the best performance of any new team.
2. RoboCup Soccer 3D Simulation
Team: Karachi Koalas was established in 2010 as a result of joint venture with the prestigious Institute of Business Administration (IBA) in Karachi. IBA was established by the Wharton Business School, University of Pennsylvania. The team qualified for the International Robot Soccer World Cup in 2010, came 16th in 2011, 10th in 2012, and 5th in 2013. This team consistently defeated the University of Oxford and the University of Paris.
Team: REEM@LASALLE (PAL Robotics Team led by Sammy Pfeiffer) 2014. The UTS team RobotAssist competed at several RoboCup@Home championships.
4. RoboCup Rescue
UTS been a contributor to RoboCup Rescue and competed in several occasions.