Research

11/02/08

Home
Research
Publications
Courses
Photo Gallery
Feedback

 

Current

1. Effects of position uncertainty in grasping 3D objects
Sponsor: National Institute of Healty (NIH)
 

This work addresses the problem of how humans compensate for the position uncertainty of objects when grasping them. One of the most challenged problems in robotics is grasping objects when their position is known imprecisely. Even though this problem is considered very difficult for robotic systems, human beings may develop strategies using prior information for handling the position uncertainty of the objects. To investigate this issue, we performed experiments in which participants reached to and lifted cylindrical objects using precision grasps. Uncertainty about the object’s location was introduced by randomly moving a binocularly viewed object using a PUMA robotic arm over sequence of 5 positions. The positions were sampled from a 2-D Gaussian distribution with a strongly oriented covariance matrix. Vision of the object was then removed using liquid crystal shutter glasses, and the robot moved the object one additional time out of view. The participant’s were then cued to rapidly reach for the object while out of view. All the object positions were sampled from the same covariance for a block of 100 trials, and data were collected for several covariance conditions by rotating orientation covariance matrix. Finger trajectories were recorded by placing 3 Optotrak sensors on top of the subjects’ thumb and index finger nails. Initially, the subject sets its thumb and index finger on a position called referred position of the trial. We tested the specific hypotheses: Does the subject build a specific strategy for grasping the object in each experiment? And is there any significant difference on the way that the subject grasps the object as the uncertainty ellipsoid varies?

2. Brain Computer Interface using Non-Invsasive Techniques (EEG, MEG)
Sponsor: VA Medical Center, University of Minnesota
 
 

Brain Computer Interface (BCI) provides communication and control to people with severe motor disabilities. It can use invasive and non-invasive methods for recording brain signals and controlling prosthetic devices. Invasive BCI uses implanted electrodes to extract the brain signal information that conveys the user’s commands. However, the implanted electrodes in the brain carry inherent risk, since they show viability of some months and they should be replaced by new electrodes. In this work, we present a new non-invasive BCI technique using Electroencephalography (EEG) for recording the brain signals. A human subject copies a pentagon for 5 minutes using an X-Y­ joystick while EEG signals are being recorded from 64 (or 32) channels. Finally, a linear summation of weighted contributions of the EEG signals is used to predict the movement trajectory.

3. Grasping a 2 Dimensional Object under shape and contact location uncertainty
Sponsor: National Institute of Healty (NIH)
 

One of the most important problems in grasping and manipulation is the selection of contact points to grasp an object. Humans have thousands of sensors and actuators to coordinate and adapt them to grasp an object. In addition, they learn how to handle uncertainty in object shape through developmental experience. Understanding the way that human grasps an object in a stochastic environment can help us to build better artificial devices (e.g., limbs). In this work, we explore the benefits that we should gain by taking uncertainty into account.

Previous

1. Next Generation Medtronic Spinal Cord Stimulator (Summer Internship)
Sponsor: Medtronic

Neurostimulation is a pain treatment that delivers low voltage electrical stimulation to the spinal cord or peripheral nerve to inhibit or block the sensation of pain. The neurostimulation system consists of stimulating lead(s), which deliver(s) electrical stimulation to the spinal cord or peripheral nerve; an extension wire, which conducts electrical pulses from the power source to the lead; and a power source, which generates the electrical pulses.

 

2. Multi Robot Trajectory Generation for Single Source Explosion Parameter Estimation
Sponsor: University of Minnesota (DTC) , National Science Foundation (NSF)



In the post-Cold War era and after the terrorist attacks on September 11, 2001, there is a growing concern that future terrorist activities may involve the use of chemical or biological weapons, such asthe attack with the nerve gas Sarin in the Tokyo subway in 1995. In addition, there is a number of unintentional chemical agent explosions, such as an accidental release from a chemical refining or storage area, as the Union Carbide release in Bhopal, India. In each situation, a quantity of hazardous material is released in the atmosphere creating dangerous or lethal conditions for people. Furthermore, the chemical agent particles may be spread by the wind and affect areas outside the immediate proximity of the initial release. Dealing with these scenarios requires that specially trained individuals are able to accurately and rapidly predict the consequences of the release, estimate its spread, and determine emergency response actions such as evacuating populated areas and/or attempting to contain the agent by spraying appropriate chemical absorbents.
In an ideal situation, this sampling task could be achieved by mobile robots. This would require that these robots are equipped with sensors for measuring the concentration of the chemical agent, registering the locations and times where measurements were recorded, and processing this information to compute accurate estimates for the parameters of the advection-diffusion equation. Robots with these capabilities will reduce the risk to the human response teams and potentially lower the financial cost of such operations since no life support equipment is necessary for a robotic emergency response team.

 

3.  Development of an Intelligent Autonomous Navigation System for Unmanned Aerial Vehicles
Members: European Aeronautics Defense and Space Systems -3SIGMA SA (EADS-3SIGMA) Technical University of Crete (TUC) , National Technical University of Athens (NTUA)
Sponsor:
Greek General Secretariat for Research and Technology


An Intelligent control system for the autonomous navigation of Unmanned  Aerial Vehicles has been developed and has been applied to the  NEARCHOS U.A.V. System. NEARCHOS is a high payload, medium range and endurance, multi-role inhabitant aerial vehicle. It can be used for military and civilian operations, such as surveillance, aerial reconnaissance, target acquisition, communication data relay, geological and oceanic applications, traffic surveillance, environmental data acquisition etc. The flight of NEARCHOS is controlled by two independent control loops. The first has as input the roll command, which is the desirable roll of the aircraft. The gyroscope, which is installed at the NEARCHOS, offers the factual roll of the vehicle. Thus, the control system compares the two angles and outputs an appropriate command to the ailerons of the aircraft. The second loop controls the pitch angle. It has two inputs, which are the desirable and the factual pitch angle and an output, which is the command to the elevators of the aircraft. 

 

4.  Preliminary Design and Development of a Turbo-Jet Engine for an Unmanned Aerial Platform
Members:  European Aeronautics Defense and Space Systems -3SIGMA SA (EADS-3SIGMA) , Technical University of Crete (TUC) , National Technical University of Athens (NTUA)  
Sponsor: Greek General Secretariat for Research and Technology

 

 

Home | Research | Publications | Courses | Photo Gallery | Feedback

This site was last updated 11/02/08