2018 REU Participants
Graduate Student Mentors
Real-world tasks are not only a series of sequential steps, but typically exhibit a combination of multiple types of constraints. These tasks pose significant challenges, as enumerating all the possible ways in which the task can be performed can lead to large representations and it is difficult to keep track of the task constraints during execution. Previously we developed an architecture that provides a compact encoding of such tasks and validated it in a single robot domain. We recently extended this architecture to address the problem of representing and executing tasks in a collaborative multi-robot setting. Firstly, the architecture allows for on-line, dynamic allocation of robots to various steps of the task. Secondly, our architecture ensures that the collaborative robot system will obey all of the task constraints. Thirdly, the proposed architecture allows for opportunistic and flexible task execution given different environmental conditions. We demonstrated the performance of our architecture on a team of two humanoid robots (a Baxter robot and a PR2 robot) performing hierarchical tasks. Further extensions of this architecture are currently being explored.
My research goal is to advance the ways robots understand their surrounding world and act within it by utilizing sensor fusion and data understanding techniques. More specifically, my focus has been on enabling greater autonomy in aerial robots to work, explore and map complex and challenging environments autonomously without the need of human intervention. The application of my research covers a wide spectrum from being deployed for inspection of infrastructure and industrial environments to helping first responders in understanding the area before going in during emergency situations e.g. earth-quakes or chemical/radioactive spillages. In technical terms, I work in two key areas namely robotic perception and path planning. In the field of robotic perception, I work on integrating and exploiting knowledge of multi-sensor systems on-board robots. They key idea here is to understand and offset the limitations of one sensor with another and make the robot choose the most beneficial sensing modality according to its environment. On the topic of path planning, I primarily work on algorithms to enable greater robot autonomy in terms of exploring and mapping challenging unknown environments efficiently without the need of constant human supervision and intervention.
Andrew (Hank) Palmer
Intent Recognition in Socially Aware Navigation: My research into intent recognition for the purposes of socially aware navigation uses a combination of computer vision and laser data to evaluate the intent of people in a given environment. The goal is to combine the intent recognition with path planning to generate more socially appropriate navigation than traditional path planning.
Adjustable Autonomy in Robotic Systems: This research is the basis of my master's thesis. It involves looking at how mmachine learning techniques can improve upon the transitions between levels of autonomy in an adjustable autonomy system. The system will examine user's patterns to anticipate appropriate autonomy levels.
Dynamic Action Spaces in Reinforcement Learning: Reinforcement learning is a powerful machine learning technique that allows the system to perform unsupervised learning. However, traditional RL suffers from large state-action spaces. This research focuses on reducing the complexity of RL state-action spaces by performing dynamic allocation of actions in a state space.
Huy Xuan Pham
Using flying robots, or Unmanned Aerial Vehicles (UAV), to assist human in difficult scenarios, such as wildfire fighting, and other natural disaster relief is very promising. They can be used to replace humans for hazardous tasks while conserving sizable operation costs in comparison with traditional methods. With a team of UAVs, we can achieve even more than that, as they provide extra maneuverability and information thanks to more capabilities to host a wide range of sensors and devices. Moreover, they can work together to improve both individual and team performances. It can potentially be a game changer in many fields. My research focuses on designing automatic control and collaborative learning algorithms for Multi Robotic Systems that allow a UAV team to coordinate to solve complex real-life problems, such as wildfire fighting or search and rescue missions.
Paulo Alexandre Regis
I am a Ph.D. candidate at the Computer Science and Engineering department at the University of Nevada, Reno. I am a member if the NetLab since 2014 when joining the doctoral program, working under the supervision of Dr. Shamik Sengupta. Previously, I obtained my Bachelor's degree in Telecommunications Engineering from Regional University of Blumenau, Brazil in 2010.
Stacey Grace Talens Cubos
Stephen "Michael" Simmons