Applications

Agriculture

As part of a team with Georgia Tech scientists and engineers, fellow(s) are working on a robotic “field doctor” designed to monitor plant health and detect insect infestation. Current focus is on bell peppers and caterpillars infestation. Two subsystems are under active research and development:

The HealthScan subsystem is designed to detect “critical” zones in the plant canopies, on both sides of the robotic vehicle as it travels down a field row, whereas the general health status of the plant would be modified as a physiological response to pest infestations.  These health changes can be measured spectrally as the variation of the “red edge” characterized by reflectance changes in the NIR range.  A ruggedized multispectral imaging system is being developed using dual CMOS cameras equipped with appropriate optical filters.

After the PestRemoval subsystem loosens pests off the plant structure, it conveys them onto a conveyor equipped with a passive brush system so as to provide a “single” layer of insects into the viewport of the PestScan subsystem.  The conveyor belt and the PestScan camera subsystem will be synchronized so as to enable the stitching of sequential images into a larger mosaic image with minimal overlaps.

Fellows involved in this area of research include:

Education

Faculty fellows from UGA’s Colleges of Education, Engineering and Arts and Sciences are collaborating on a project called RoboTube for Teacher Education. RoboTube is a video-sharing website tailored to the use of robotics in K-12 education. The project is designed to provide pre-service and in-service teachers the opportunity to learn how to use robotics technology in their current and future classrooms. Research shows that robotics technology has the potential to significantly help teachers facilitate student engagement and active learning. The ultimate goal of the project is to positively influence not only teachers’ learning from teacher education courses but also student learning in Georgia and beyond. In another line of research, faculty fellows are investigating emotions and emotion regulation in students by involving various technologies to improve learning and performance. A project in this area is investigating if students’ empathy toward robots is related to the quality of the robots they build.

When analyzing student “bipedal” projects from robotics courses, fellows observed that the logical-mathematical skills (required for computer programming of robots), and body-centered mechanical empathy skills (required for programming flowing gait solutions for robots) were 2 independent skill sets and that one set of skills did not necessarily imply the existence of the other for typical engineering students. Faculty fellows are collaborating toward developing an educational framework to help study this area further.

Fellows involved in this area of research include:

Entertainment

Theatre robotics involves applying robotic technologies to the arts, in particular theatre and performance art, and conversely the application of theatrical expertise (writing, acting, directing) to robotic design. The key objectives of research in this area are: (1) to create engaging, highly entertaining performances with robots and live actors (in effect robotic puppetry); (2) to create high-level tools that allow actors, directors and animators without engineering or programming skills to animate robots, and to synchronize the robots’ behavior to other theatrical elements such as sound, video and stage lighting; (3) to develop technologies and techniques — both in hardware and software — that allow robots to interact and improvise effectively with live performers in real time; (4) to apply the results of this research in theatre robotics to more general areas of robotic science, in particular interactivity and social robotics.

As part of a project called “Commedia Robotica”, which involves robotic actors in the Commedia dell’Arte form of theater, the main research goal is to remove the human operator from the loop of interactions between a human actor and a humanoid robot.

As part of this umbrella project, faculty fellows are investigating several research thrusts. These include:

  • Real-time (or fast-enough) capture and characterization of meaningful gestures of the human actor from gesture cameras such as Microsoft’s Kinect or SoftKinetic’s DS311.
  • Development of a “Character Engine” (young/old, happy/sad) at the servo motors PID control level for the humanoid robot.
  • Software development for a Theatre Planner/Learner tool using production systems tools such as CLIPS or SOAR.
  • Hidden Markov models for “Theatrical Flow” to go from act to act within a performance session.
  • Development of a “human spectator” assessment tool using Clickers technology so that the humanoid robot can “learn” from its own performance using intelligent machine learning techniques.

Fellows involved in this area of research include:

Healthcare

Medical robotics is a multi-disciplinary research area that combines mechanical engineering, electrical engineering, computer science and medicine. Faculty fellows are engaged in developing technology that will overcome surgical limitations, improve diagnosis and rehabilitation, and improve the ability of clinicians to perform those tasks. The primary research foci include the integration of mechatronics and surgical robots with MRI technology, image-guided minimally invasive surgery, as well as smart diagnostic and surgical devices. Fellows are actively collaborating with engineers, computer scientists, industrial companies, and hospitals to bring medical robotic technology into everyday clinical environments.

Fellows have been previously involved in designing and prototyping a broad range of smart surgical robots and clinical devices, most of which have been applied in numerous clinical trials in patients.

Fellows involved in this area of research include:

Manufacturing

Research in this crucial application area focuses on cycle time analysis in robotic cells. In particular, faculty fellows are interested in investigating the sequencing and scheduling of robot movement in an environment with many possible alternative routes. Typical motion of an industrial robot is load, move, and unload. Studying a throughput of a cell is partially dependent upon sequence of robot moves as well as the processing time of parts on the production machines. Research involves both cyclic and non-cyclic production that aims at minimizing time for production.

Fellows involved in this area of research include:

  • John Mativo