Padır's Corner

There is no such thing as public opinion. There is only published opinion. -Winston Churchill

Robotics Seminar: Stefanie Tellex, Brown University

Learning Models of Language, Action and Perception for Human-Robot Collaboration
Stefanie Tellex, Brown University
Time: 8 February 17, Wednesday, 11am
Place: Richards Hall 458
Abstract: Robots can act as a force multiplier for people, whether a robot assisting an astronaut with a repair on the International Space station, a UAV taking flight over our cities, or an autonomous vehicle driving through our streets.  To achieve complex tasks, it is essential for robots to move beyond merely interacting with people and toward collaboration, so that one person can easily and flexibly work with many autonomous robots.  The aim of my research program is to create autonomous robots that collaborate with people to meet their needs by learning decision-theoretic models for communication, action, and perception.  Communication for collaboration requires models of language that map between sentences and aspects of the external world. My work enables a robot to learn compositional models for word meanings that allow a robot to explicitly reason and communicate about its own uncertainty, increasing the speed and accuracy of human-robot communication.  Action for collaboration requires models that match how people think and talk, because people communicate about all aspects of a robot’s behavior, from low-level motion preferences (e.g., “Please fly up a few feet”) to high-level requests (e.g., “Please inspect the building”).  I am creating new methods for learning how to plan in very large, uncertain state-action spaces by using hierarchical abstraction.  Perception for collaboration requires the robot to detect, localize, and manipulate the objects in its environment that are most important to its human collaborator.  I am creating new methods for autonomously acquiring perceptual models in situ so the robot can perceive the objects most relevant to the human’s goals. My unified decision-theoretic framework supports data-driven training and robust, feedback-driven human-robot collaboration.
Bio: Stefanie Tellex is an Assistant Professor of Computer Science and Assistant Professor of Engineering at Brown University.  Her group, the Humans To Robots Lab, creates robots that seamlessly collaborate with people to meet their needs using language, gesture, and probabilistic inference, aiming to empower every person with a collaborative robot.  She completed her Ph.D. at the MIT Media Lab in 2010, where she developed models for the meanings of spatial prepositions and motion verbs.  Her postdoctoral work at MIT CSAIL focused on creating robots that understand natural language.  She has published at SIGIR, HRI, RSS, AAAI, IROS, ICAPs and ICMI, winning Best Student Paper at SIGIR and ICMI, Best Paper at RSS, and an award from the CCC Blue Sky Ideas Initiative.  Her awards include being named one of IEEE Spectrum’s AI’s 10 to Watch in 2013, the Richard B. Salomon Faculty Research Award at Brown University, a DARPA Young Faculty Award in 2015, and a 2016 Sloan Research Fellowship.  Her work has been featured in the press on National Public Radio, MIT Technology Review, Wired UK and the Smithsonian.  She was named one of Wired UK’s Women Who Changed Science In 2015 and listed as one of MIT Technology Review’s Ten Breakthrough Technologies in 2016.

Robotics Seminar: Professor Masayuki Inaba, University of Tokyo

We are pleased to host Professor Masayuki Inaba and his colleagues at Northeastern’s Robotics Collaborative. Professor Inaba leads JSK Robotics Lab at the University of Tokyo which is the home of numerous humanoid robots including Kengoro and JAXON. Here are the details of Professor Inaba’s seminar talk.

Recent Robotics Activities and Background at JSK Lab, University of Tokyo
Professor Masayuki Inaba

Time: 13 December 2016, 2pm
Place: Richards Hall 300

The talk will introduce recent ongoing activities of the research and development at the JSK Robotics Lab, University of Tokyo which include our DRC humanoid, JAXON, and musculoskeletal humanoid, Kengoro. At JSK Lab (http://www.jsk.t.u-tokyo.ac.jp/research.html). We are working on several research topics on system architecture and integration for task execution, continuous perception with attention control, variable robots for system abstraction, hardware platform like HRP2 and PR2 for general purpose system, software platform with lisp-based programming environment, asynchronous multi-component envrironment of RTM-ROS environment with continuous integration tools, specialization for industry collaboration and hardware challenges for robot vision, whole-body tactile and flesh sensors, flexible and redundancy with complexity of muscloskeletal, high-power drives with heat transfer and control, flying and hovering in space, and so on.

Masayuki Inaba is a professor of Department of Creative Informatics and Department of Mechano-Informatics of the graduate school of information science and technology of The University of Tokyo. He received B.S of Mechanical Engineering in 1981, Dr. of Engineering from The University of Tokyo in 1986. He was appointed as a lecturer in 1986, an associate professor in 1989, and a professor in 2000 at The University of Tokyo. His research interests include key technologies of robotic systems and their research infrastructure to keep continuous development for advance robotics.

Professors Dennis Hong (UCLA), Masayuki Inaba (University of Tokyo) and Taskin Padir. Photo taken on 24 April 2014 at the JSK Robotics Lab, University of Tokyo.

Project ATHENA SRC

We received very exciting news yesterday… NASA awarded our team a Valkyrie humanoid robot. We are looking forward to collaborating with NASA, our colleagues at MIT and Space Robotics Challenge competitors to advance the autonomy on Valkyrie. Here is a brief summary of our project acknowledging our entire team:

Accessible Testing on Humanoid-Robot-R5 and Evaluation of NASA Administered (ATHENA) Space Robotics Challenge

Taskin Padir, Associate Professor, Electrical and Computer Engineering, Northeastern University
Robert Platt, Assistant Professor, College of Computer and Information Science, Northeastern University
Holly Yanco, Professor, Computer Science Department, University of Massachusetts, Lowell

Our overarching goal in this basic and applied research and technology development effort is to advance humanoid robot autonomy for the success of future space missions. We will achieve this goal by (1) establishing a tight collaborative environment among our institutions (Northeastern University (NEU) and the University of Massachusetts Lowell (UML)) and NASA’s Johnson Space Center, (2) leveraging our collective DARPA Robotics Challenge (DRC) experience in humanoid robot control, mobility, manipulation, perception, and operator interfaces, (3) developing a systematic model-based task validation methodology for the Space Robotics Challenge (SRC) tasks, (4) implementing novel perception based grasping and human-robot interaction techniques, (5) providing access to collaborative testing facilities for the SRC competition teams, and (6) making the developed software available to the humanoid robotics community. Successful completion of this project will not only progress the technological readiness of humanoid robots for practical applications but also nurture a community of competitors and collaborators to enhance the outcomes of the SRC to be administered by NASA in 2016. We propose to unify our team’s complementary expertise in robot control (Padir), grasping (Platt), and human-robot interaction (Yanco) to advance the autonomy of NASA’s R5. Since August 2012, Padir has been co-leading the WPI DRC Team, the only Track C team that participated in the DRC Finals with an Atlas humanoid robot. Platt participated in Team TRACLabs’ DRC entry to enhance Atlas robot’s autonomous manipulation capabilities. Yanco led the DARPA-funded study of human-robot interaction (HRI) for the DRC, at both the Trials and the Finals. Our team is unique in terms of facilities and capabilities for hosting NASA’s R5 at the New England Robotics Validation and Experimentation (NERVE) Center at UMass Lowell in Lowell Massachusetts, less than 30 miles from Boston. At 10,000 square feet, the NERVE Center is the largest indoor robot test facility in New England.

CUCZ6ZCWEAAqHB9

RIVeR Lab moves to Northeastern University…

Since its founding in 2010, Robotics and Intelligent Vehicles Research (RIVeR) Laboratory has been the home of many researchers from PhD students to high-school interns who worked on a variety of projects from assistive robotics to robots for disaster response. As of September 1, 2015, we announce that RIVeR Lab moves to Northeastern University. We are now recruiting researchers at all levels. For more information contact Professor Padir at t.padir@northeastern. edu

 

Northeastern University

Recruiting MQP-VINE Students

WPI’s Robotics and Intelligent Vehicles Research (RIVeR) Laboratory (robot.neu.edu) is leading a new initiative called Major Qualifying Project-Vertically Integrated Experience (MQP-VINE, in short). Within the scope of an MQP-VINE, students will have an opportunity to begin their MQP experience as early as their sophomore year by getting involved in our cross-disciplinary sponsored research projects on assistive robotics, autonomous exploration rovers, aerial vehicles, human-in-the-loop robot control, and cyber physical systems. In its full implementation, we envision that students will gain theoretical knowledge and practical skills in robotics engineering by completing their MQP in functional multidisciplinary project teams of sophomores, juniors, and seniors on well-scoped yet challenging projects.

In a nutshell, here’s how it works.
– If you are a rising sophomore now, you will complete your MQP over three years by registering 1/3 units each year.
– If you are a rising junior, you will complete your MQP over two years by registering 1/3+2/3 units.
– It is expected that you will spread your units over at least two terms each year.
– It is expected that you will commit 9-11 hours (on average) per week for each 1/6 unit that you register.
– It is expected that you will “stay with problems” for the duration of the project regardless of the units you register.

At this time, we are seeking applications from rising sophomores and rising juniors majoring in RBE, ECE, CS, and ME programs. Double majors are welcome to apply. We expect to recruit up to three sophomores and three juniors this year.

Please send the following information to tpadir@wpi.edu if you are interested in joining the MQP-VINE in RIVeR Lab.

– A resume including your name, contact information, academic standing, project work, skills, and other relevant information.
– An unofficial copy of your transcript from Bannerweb.
– A short essays (200 words or less) on your research interests, goals, and preparation.

All relevant questions can be directed to Professor Padir, tpadir@wpi.edu.

Recruiting for 2015-16 MQPs

RIVeR Lab is planning to advise/sponsor the following MQPs (Major Qualifying Projects) for the 2015-16 academic year. If you are interested in any of these projects please contact Professor Padir, tpadir@wpi.edu.

A. Assistive Robots for Improving Quality of Life for Older Adults 

(Please note, we will recruit two project teams on this topic.)

RIVeR Lab is working with local assisted living facilities and nursing homes to develop assistive robots to foster aging in place. With support from the National Science Foundation, we will advise two projects on designing assistive robots that can operate in home environments and perform activities of daily living. One project will develop robust software for WPI’s personal assistant robot FRASIER (Fostering Resilient Aging with Self-Efficacy and Independence Enabling Robot). The second project will utilize rapid prototyping techniques to realize CHEER (Co-robots for in-Home Evaluation of Environmental Risks) to prevent falls. Project teams will also work towards enhancing our lab’s connections with non-profit organizations and government agencies and businesses. Both projects will be validated within WPI’s @Home testbed.

B. Robotics for Advanced Response to Epidemics (RARE)

(Please note, we will recruit multiple project teams on this topic.)

Project team will work closely with our researchers to design and validate new gadgets and techniques within our emergency treatment unit testbed. The overarching goal of this NSF-funded project is to develop new technologies for preventing, detecting and responding to current and future epidemics (such as Ebola). Our current capabilities include telepresence robots, technologies for remote patient monitoring, bed-side technologies and decontamination robots.

C. Accelerating Mammalian Embryo Research Using Co-Robots

This project will be completed in close collaboration with the University of Massachusetts Medical School, Department of Cell and Developmental Biology. The study of early mammalian development requires the analysis of gene expression at the whole organism level. This allows for the temporal and spatial visualization of the patterns of gene expression that control embryonic development. The foremost technique used in these studies is wholemount in situ hybridization. This technique is highly laborious and tedious consisting of approximately 50 steps that vary in time of incubation, and temperature with the whole technique spanning three days. The overarching goal in this collaborative project is to enhance the experimental research outcomes, minimize human errors, and accelerate the training of next generation developmental biology researchers by iteratively designing, prototyping and validating a new class of rapidly engineered co-robots that can work side-by-side with developmental biology researchers.

D. WALRUS Rover Autonomy and Payload Development

The project team will advance the capabilities of WALRUS Rover: http://robot.neu.edu/walrus/

We continue to add more project ideas here…

2015-16 IQP Topics

RIVeR Lab is planning to advise the following IQPs (Interactive Qualifying Projects) for the 2015-16 academic year. If you are interested in any of these projects please contact Professor Padir, tpadir@wpi.edu.

2016 WPI RoboEthics Symposium

This IQP is on planning the 2016 WPI Roboethics Symposium. DARPA Robotics Challenge Finals will feature 25 teams from all around the world. Even though the challenge is on designing robot hardware and software for disaster response, what are the ethical issues around the dual use of technology? FAA recently selected 6 aerial vehicle test sites in the United States. Are there privacy and security issues with personal use of drones? In 2014, the SmartAmerica Challenge organized by Presidential Innovation Fellows featured several projects on healthcare and patient monitoring. What are the ethical issues surrounding smart technologies in elder care. In 2010, we hosted the 1st WPI RoboEthics Symposium (http://www.wpi.edu/Pubs/E-project/Available/E-project-042510-210828/). It is now time to continue the dialogue.

Creating a Research Roadmap for Robotics for Advanced Response to Epidemics (RARE)

Motivated by the fact that the 2014 Ebola outbreak is the largest in history, there is a pressing need to understand how to improve clinical care, logistical support, and situational awareness during epidemics. This IQP will build on our lessons learned from the Brainstorming Workshops on Safety Robotics for Ebola Workers held on Nov. 7, 2014, and 2015 RSS Workshop on Robotics for Advanced Response to Epidemics (RARE) to be held in July 17, 2015 to create a research and development roadmap for RARE.

User Experience Design for FRASIER

It is projected that in 2030, there will be 4 people for each person over the age of 65. Among these four people, one will be a child, one will be sick and one will be at a distant geographical location. This implies that the ratio of younger adults as caregivers to older adults as individuals in need of care will be 1-to-1 in 2030.  Robotics technologies, the integration of sensing, computation and actuation in the physical world, can be used to transform the capabilities of a person with a disability in performing the activities of daily living. Robots will not replace a human caregiver in near future, however they can provide an extended independent living for older adults or individuals with physical and mental disabilities as personal assistants or companions, and hence, improve the quality of life for humans. This IQP will investigate the acceptability and user perception for new robot technologies for assisting persons with general and age-related disabilities and create a user experience design for WPI’s personal assistance robot, FRASIER (Fostering Resilient Aging with Self-efficacy and Independence Enabling Robot).

STEM Saturdays @WPI

It was great to share our vision for realizing Cyber-physical systems for Advanced Response to Epidemics (CARE) with the STEM Saturdays participants, this morning. Thank you parents for all the engaging discussions as your middle-schoolers worked on their projects. We also thank WPI’s Office of Multicultural Affairs for their invitation.

IMG_0285-0.JPG