Human-Robot Teaming

Mentor: Dr. Julie A. Adams
Website: https://research.engr.oregonstate.edu/hmtl/

This project seeks to develop new methods for adapting interaction between humans and robots based on the human's workload and awareness. The human’s physiological data is analyzed to detect nominal and potentially detrimental workload conditions,. The robot can adapt its interaction method, autonomy level, or assigned tasks based on the human’s state. REU students will work to develop new algorithms for for adapting the robot's interactive capabilities based on the workload detection. The project will require algorithm design, implementation and software testing. Further, REU students will evaluate the algorithms with actual human-robot teams.

Re-engineering the Musculoskeletal System Using Novel Robotics-inspired Passive Orthopedic Implants

Mentor: Dr. Ravi Balasubramanian
Website: https://web.engr.oregonstate.edu/~balasubr/

Current reconstructive orthopedic surgeries use sutures to attach muscles and tendons. However, this leads to poor surgical outcomes because of the suture’s limited ability to transmit the muscle’s forces and movement to the tendons. It is expected that using passive implants, such as pulleys and rods, to surgically construct mechanisms in situ using the existing biological tendons will significantly improve post-surgery function (when compared to using sutures) and lead to the development of new surgical procedures. 

Example projects:

Robotic Crop Manipulation
Mentor: Dr. Joe Davidson and Dr. Cindy Grimm
Website: https://research.engr.oregonstate.edu/davidsonjr/ and https://cindygr.github.io/

Robots are needed to help with the production of specialty crops such as fresh market fruits and vegetables. However, specialty crop cultivation systems are challenging, unstructured environments. Some of the challenges include uneven terrain, variable lighting conditions, occlusions, clutter, and delicate products that can be easily damaged. Another constraint is the seasonal nature of agriculture that limits data collection in the field (the fruit may only be ripe for a few days each year!). Our work aims to advance manipulation in unstructured environments like specialty crop systems through a combination of tactile sensing, soft robotics, and learning from physical twins. This project will involve field testing at commercial farms around the Willamette Valley.

Example projects include:

  • Building grippers for picking fruit
  • Developing physical testbeds
  • Creating simulation environments to study crop manipulation

Soft Robots for Exploration and Manipulation

Mentor: Dr. Brian Do
Website: https://doroboticslab.org/

Soft robots can take on a variety of form factors and enable new designs not possible with existing robots. My work aims to design new types of robots that can change their shape and mechanical properties to enable new types of interactions between robots and the world around them. For example, soft robots capable of growth, often referred to as “vine” robots, can be used for grasping objects like an elephant trunk or exploring rubble like a snake. The REU student would be involved with the design, modeling, and/or fabrication of new types of soft robots.

Example projects include:

  • Building growing soft robot arms
  • Developing new soft grippers
  • Building soft robots for exploring underwater or in other environments

Teaching Humanoid Robots to Do Things
Mentor: Dr. Alan Fern
Website: https://mime.engineering.oregonstate.edu/research/drl/

Humanoid robots have the physical capability to perform much of the physical work that is currently done by humans. Currently, however, the algorithms for intelligently controlling humanoids are unable to address the vast majority of human work tasks. The Dynamic Robotics and AI Lab (DRAIL) at OSU studies machine-learning approaches to advancing humanoid control, starting from low-level whole-body motion to higher-level task planning. The long-term objective is to enable humans to instruct humanoids, via natural language, to reliably carryout tasks involving basic physical common-sense and skills. This involves working on a tight integration of reinforcement learning, computer vision, language modeling, and high-level planning.

Example Projects:

  • Developing interfaces, such as VR, to teach humanoids physical skills.
  • Training humanoids in simulation, via reinforcement learning, to autonomously learn new skills.
  • Studying approaches to improve transfer of humanoid controllers from simulation to the real world.  
  • Designing simulation and real-world humanoid benchmarks and evaluating controllers on the benchmarks.

Robots for Health Promotion
Mentor: Dr. Naomi Fitter
Website: https://osusharelab.com/research/

Compared to other types of interactive technologies, robots possess a unique ability to motivate people because people tend to perceive them as a "social other," rather than a tool or device. The OSU SHARE Lab studies applications of robots in health-promoting scenarios, including mobility interventions and physical activity encouragement. For example, children are becoming more sedentary over time; we are curious about the role robots can play in encouraging physical exploration and play for children with typical development and children with motor disabilities. Older adult wellness is a huge topic of interest and study as the need for elder care exceeds the current capacity of the healthcare system; we are researching ways that robots could help to support physical, cognitive, and social wellness in this population.

Example projects:

  • Supporting the design and deployment of robotic systems for interventions with people
  • Designing and evaluating robot behaviors
  • Gathering and analyzing human user data

Autonomy for Underwater Vehicles Exploring and Manipulating Ocean Environments
Mentor: Dr. Geoff Hollinger
Website: http://research.engr.oregonstate.edu/rdml/

There are many ocean environments which are unsuitable for crewed research vessels, either because they are too dangerous (e.g., under an ice shelf or in the deep ocean), or require too many resources to be effective. We seek to design a new generation of Resident Autonomous Vehicles (RAVs) that can stay in place for long periods of time to measure ocean dynamics and perform maintenance/inspection in extreme environments. This research project involves the programming, hardware design, and construction of robust autonomous underwater systems to explore the ocean in Greenland, Antarctica, and in remote ocean basins. The REU student will join an interdisciplinary team of researchers from mechanical engineering, computer science, and oceanography to assist in building and programming the RAVs currently being designed at Oregon State University.

Example projects:

  • Designing algorithms for robust underwater manipulation and grasping
  • Optimization and testing of autonomous underwater docking capabilities
  • Programming autonomous vehicles to operate with minimal operator control
  • Learning from human operators to coordinate autonomous marine vehicles
  • Designing multi-robot mothership-passenger systems for remote under-ice environments

Human-Robot Social Interaction
Mentor: Dr. Bill Smart
Website: https://engineering.oregonstate.edu/people/bill-smart

As robots enter our daily lives, they will have to learn the social rules that humans follow.  How close can you stand to someone?  When can you interrupt to ask a question without being annoying?  Who gets priority when two people are trying to go different ways through the same doorway?  Although we do these things every day without thinking about them, they can be challenging for robots to get right.  This project will involve designing, implementing, and evaluating social interactions between a Quori robot and people.  We’ll be building on prior work in both social psychology and social robotics, and evaluating our work in real-world settings on the OSU campus.

Human-Robot Teaming for Field Science Data Collection
Mentor: Dr. Cristina Wilson
Website: https://www.radlab.us/

Robots are used in Earth and planetary science missions as mobile sensor suites, taking low-level commands from humans to execute the navigation, sensing and sampling, while human experts bear the full burden of integrating and interpreting data for future data collection decision making. The goal of this NASA-funded project is to develop new human-robot teaming workflows that allow robots to take on increased responsibility in collaborative exploration decision making, thereby freeing up the expert to engage in the type of abstract hypothetical thinking that the human mind excels at.

The REU student(s) will help with user studies evaluating how scientists respond to suggestions from a robot about where to collect data next. There will be opportunity to conduct studies in simulation and during a planetary analogue science mission at White Sands National Monument during the summer.