Richard M. Voyles

Enlisting rangers and scouts for reconnaissance and surveillance

Reconnaissance and surveillance are important activities for both military and civilian organizations, for hostage and survivor rescue, drug raids, response to chemical or toxic waste spills etc. We have developed a distributed heterogeneous robotic team that is based mainly on a miniature robotic system. Because some operations require covert action, most of the robots are extremely small. This also allows them to be easily transported and allows for a greater number to be brought into use for a single operation. This makes them expendable without jeopardizing the overall mission. We call these small robots scouts. Their individual components must all be exceedingly small, and their overall design must make maximum use of all available space. They must make efficient use of resources (e.g., batteries). We meet these challenges with an innovative design and creative use of additional support. We team the scouts with larger ranger robots, which can transport the scouts over distances of several kilometers, deploy them rapidly over a large area, coordinate their behavior, and collect and present the resulting data. We present the scouts and rangers, discuss their capabilities along with the associated software, and describe demonstrations conducted to test the innovative aspects of the system. We also discuss related work, analyze our results, and draw conclusions.

Human activity recognition using multi-features and multiple kernel learning

This paper presents two sets of features, shape representation and kinematic structure, for human activity recognition using a sequence of RGB-D images. The shape features are extracted using the depth information in the frequency domain via spherical harmonics representation. The other features include the motion of the 3D joint positions (i.e. the end points of the distal limb segments) in the human body. Both sets of features are fused using the Multiple Kernel Learning (MKL) technique at the kernel level for human activity recognition. Our experiments on three publicly available datasets demonstrate that the proposed features are robust for human activity recognition and particularly when there are similarities among the actions.

FPGA implementation of closed-loop control system for small-scale robot

Small robots can be beneficial to important applications such as civilian search and rescue and military surveillance, but their limited resources constrain their functionality and performance. To address this, a reconfigurable technique based on field-programmable gate arrays (FPGAs) may be applied, which has the potential for greater functionality and higher performance, but with smaller volume and lower power dissipation. This project investigates an FPGA-based PID motion control system for small, self-adaptive systems. For one channel control, parallel and serial architectures for the PID control algorithm are designed and implemented. Based on these one-channel designs, four architectures for multiple-channel control are proposed and two channel-level serial (CLS) architectures are designed and implemented. Functional correctness of all the designs was verified in motor control experiments, and area, speed, and power consumption were analyzed. The tradeoffs between the different designs are discussed in terms of area, power consumption, and execution time with respect to number of channels, sampling rate, and control clock frequency. The data gathered in this paper will be leveraged in future work to dynamically adapt the robot at run time

The Shape From Motion Approach to Rapid and Precise Force/Torque Sensor Calibration

We present a new technique for multi-axis force/torque sensor calibration called shape from motion. The novel aspect of this technique is that it does not require explicit knowledge of the redundant applied load vectors, yet it retains the noise rejection of a highly redundant data set and the rigor of least squares. The result is a much faster, slightly more accurate calibration procedure. A constant-magnitude force (produced by a mass in a gravity field) is randomly moved through the sensing space while raw data is continuously gathered. Using only the raw sensor signals, the motion of the force vector (the motion) and the calibration matrix (the shape) are simultaneously extracted by singular value decomposition. We have applied this technique to several types offorce/torque sensors and present experimental results for a 2-DOF fingertip and a 6-DOF wrist sensor with comparisons to the standard least squares approach.

TerminatorBot: a novel robot with dual-use mechanism for locomotion and manipulation

As part of a massively distributed heterogeneous system, TerminatorBot, a novel, centimeter-scale crawling robot, has been developed to address applications in surveillance, search-and-rescue, and planetary exploration. Its two three-degree-of-freedom arms, which stow inside the cylindrical body for ballistic deployment and protected transport, comprise a dual-use mechanism for manipulation and locomotion. The intended applications require a small, rugged, and lightweight robot, hence, the desire for dual use. TerminatorBots unique mechanism provides mobility and fine manipulation on a scale currently unavailable. To facilitate manipulation, we have also developed a specialized force/torque sensor. This new sensor design has a biased distribution of flexures, which equalizes force and torque sensitivities at the operational point. This work describes the mechanism and design of TerminatorBot, the specialized force/torque sensor, and the mechanism-specific gaits.

Real-Time Detection of Camera Tampering

This paper presents a novel technique for camera tampering detection. It is implemented in real-time and was developed for use in surveillance and security applications. This method identifies camera tampering by detecting large differences between older frames of video and more recent frames. A buffer of incoming video frames is kept and three different measures of image dissimilarity are used to compare the frames. After normalization, a set of conditions is tested to decide if camera tampering has occurred. The effects of adjusting the internal parameters of the algorithm are examined. The performance of this method is shown to be extremely favorable in real-world settings.

Calibration of Multi-Axis MEMS Force Sensors Using the Shape-From-Motion Method

Precise calibration of multi-axis microelectromechanical systems (MEMS) force sensors is difficult for several reasons, including the need to apply many known force vectors at precise orientations at the micro- and nanoNewton (nN) force scales, and the risk of damaging the small, fragile microdevices. To tackle these challenges, this paper introduces the shape-from-motion calibration method. A new design of a two-axis MEMS capacitive force sensor with high linearity and nN resolutions is presented. Structural-electrostatic coupled-field simulations are conducted in order to optimize the sensor design. The designed sensor is calibrated with the shape-from-motion method, the least-squares method as well as the gravity-based method for comparison purposes. Calibration results demonstrate that the shape-from-motion method provides a rapid, practical, and accurate technique for calibrating multi-axis MEMS sensors

TerminatorBot: a robot with dual-use arms for manipulation and locomotion

A novel, miniature robot designed to use its two arms for both manipulation and locomotion is described. Intended for military and civilian surveillance and search-and-rescue applications, the robot must be small, rugged, and lightweight, hence the desire for dual-use. The robot consists of two, three-degree-of-freedom arms that can stow completely inside the 75 mm diameter cylindrical body for ballistic deployment. This paper describes the mechanism and design motivation as well as two novel locomotion gaits and a third conventional gait.

Hexrotor UAV platform enabling dextrous interaction with structures — Preliminary work

A force closure grasp is a term from dexterous manipulation that indicates a grasp that can resist any applied wrench, or force-torque. Force closure grasps are desirable because they can completely immobilize an object or impart an arbitrary wrench to an object. With fixed-base manipulators, determining the degree of force closure of a manipulation system is simplified to determining the degree of force closure of the gripper or end effector. In mobile manipulation, the manipulator base is not fixed to the ground so determining the set of wrenches that can be resisted is not strictly limited to the capabilities of the end effector. But due to the large differences in mass of the mobile base and end effector, it is generally safe to assume the degree of force closure is limited by the end effector and not by the ability of the mobile base to remain motionless. As aerial mobile manipulation has started to become an active area of research, the concept of force closure of the entire manipulation system needs to be considered. Conventional aerial platforms are not able to resist an arbitrary wrench so an end effector carried by such a vehicle will not be able to exhibit force closure. This is true because even current quad rotors lack both the number of degrees of freedom but also independence of the degrees of freedom due to the fact that the force vectors are all parallel. We have developed a hex-rotor system with six independent degrees of freedom providing force closure for a dexterous aerial vehicle for mobile manipulation tasks. The ability of the aerial mobile base to exert an arbitrary wrench coupled with a low degree of freedom manipulator will allow for an agile aerial mobile manipulator with true force closure.

Hexrotor UAV platform enabling dextrous interaction with structures-flight test

In this paper, we present the development of Dexterous Hexrotor, a hexrotor UAV platform with canted thrusters, enabling dexterous interaction with structures. Aerial mobile manipulation is an emerging niche in the field of mobile manipulation. Although there has been a fair amount of study of free-flying satellites with graspers and yielded impressive results, it is hampered a lack of appropriate testbeds for aerial mobile manipulation. Typical helicopters or quadrotors cannot instantaneously resist or apply an arbitrary force in the plane perpendicular to the rotor axis. They lack of force closure (a term from the dexterous manipulation community), which makes them inadequate for complex mobile manipulation tasks. The Collaborative Mechatronics Lab is addressing this instrumentation gap with the development of Dexterous Hexrotor to eventually host a low-cost, lightweight Stewart-Gough platform that can be combined as a macro/micro mobile manipulation system. Based on the concept of force closure, the new type of 6 DoFs hexrotor UAV provides the unique capability of being able to resist any applied wrench, or generalized force-torque. In this paper, we describe how Dexterous Hexrotor provides this important capability. We also describe the flight test which Dexterous Hexrotor is exhibiting holonomic behavior.