Jonathan Kofman

Teleoperation of a robot manipulator using a vision-based human-robot interface

Remote teleoperation of a robot manipulator by a human operator is often necessary in unstructured dynamic environments when human presence at the robot site is undesirable. Mechanical and other contacting interfaces used in teleoperation require unnatural human motions for object manipulation tasks or they may hinder human motion. Previous vision-based approaches have used only a few degrees of freedom for hand motion and have required hand motions that are unnatural for object manipulation tasks. This paper presents a noncontacting vision-based method of robot teleoperation that allows a human operator to communicate simultaneous six-degree-of-freedom motion tasks to a robot manipulator by having the operator perform the three-dimensional human hand-arm motion that would naturally be used to complete an object manipulation task. A vision-based human-robot interface is used for communication of human motion to the robot and for feedback of the robot motion and environment to the human operator. Teleoperation under operator position control was performed with high accuracy in object placement on a target. Semi-autonomous traded and shared control using robot-vision guidance aided in achieving a more accurate positioning and orientation of the end-effector for object gripping tasks.

Review of fall risk assessment in geriatric populations using inertial sensors

BackgroundFalls are a prevalent issue in the geriatric population and can result in damaging physical and psychological consequences. Fall risk assessment can provide information to enable appropriate interventions for those at risk of falling. Wearable inertial-sensor-based systems can provide quantitative measures indicative of fall risk in the geriatric population.MethodsForty studies that used inertial sensors to evaluate geriatric fall risk were reviewed and pertinent methodological features were extracted; including, sensor placement, derived parameters used to assess fall risk, fall risk classification method, and fall risk classification model outcomes.ResultsInertial sensors were placed only on the lower back in the majority of papers (65%). One hundred and thirty distinct variables were assessed, which were categorized as position and angle (7.7%), angular velocity (11.5%), linear acceleration (20%), spatial (3.8%), temporal (23.1%), energy (3.8%), frequency (15.4%), and other (14.6%). Fallers were classified using retrospective fall history (30%), prospective fall occurrence (15%), and clinical assessment (32.5%), with 22.5% using a combination of retrospective fall occurrence and clinical assessments. Half of the studies derived models for fall risk prediction, which reached high levels of accuracy (62-100%), specificity (35-100%), and sensitivity (55-99%).ConclusionsInertial sensors are promising sensors for fall risk assessment. Future studies should identify fallers using prospective techniques and focus on determining the most promising sensor sites, in conjunction with determination of optimally predictive variables. Further research should also attempt to link predictive variables to specific fall risk factors and investigate disease populations that are at high risk of falls.

Comparison of linear and nonlinear calibration methods for phase-measuring profilometry

In phase-shifting-based fringe-projection surface-geometry measurement, phase unwrapping techniques produce a continuous phase distribution that contains the height information of the 3-D object surface. Mapping of the phase distribution to the height of the object has often involved complex derivations of the nonlinear relationship. In this paper, the phase-to-height mapping is formulated using both linear and nonlinear equations, the latter through a simple geometrical derivation. Furthermore, the measurement accuracies of the linear and nonlinear calibrations are compared using measurement simulations where noise is included at the calibration stage only, and where noise is introduced at both the calibration and measurement stages. Measurement accuracies for the linear and nonlinear calibration methods are also compared, based on real-system measurements. From the real-system measurements, the accuracy of the linear calibration was similar to the nonlinear calibration method at the lower range of depth. At the higher range of depth, however, the nonlinear calibration method had considerably higher accuracy. It seems that as the object approaches the projector and camera for the higher range of depth, the assumption of linearity based on small divergence of light from the projector becomes less valid.

Two-step triangular-pattern phase-shifting method for three-dimensional object-shape measurement

Existing phase-shifting measurement methods involve processing of three acquired images or computation of functions that require more complex processing than linear functions. This paper presents a novel two-step triangular-pattern phase-shifting method of 3-D object-shape measurement that combines advantages of earlier techniques. The method requires only two image-acquisition steps to capture two images, and involves projecting linear grayscale-intensity triangular patterns that require simpler computation of the intensity ratio than methods that use sinusoidal patterns. A triangular intensity-ratio distribution is computed from two captured phase-shifted triangular-pattern images. An intensity ratio-to-height conversion algorithm, based on traditional phase-to-height conversion in the sinusoidal-pattern phase-shifting method, is used to reconstruct the object 3-D surface geometry. A smaller pitch of the triangular pattern resulted in higher measurement accuracy; however, an optimal pitch was found, below which intensity-ratio unwrapping failure may occur. Measurement error varied cyclically with depth and may partly be due to projector gamma nonlinearity and image defocus. The use of only two linear triangular patterns in the proposed method has the advantage of less processing than current methods that process three images, or methods that process more complex functions than the intensity ratio. This would be useful for high speed or real-time 3-D object-shape measurement.

Multiple-step triangular-pattern phase shifting and the influence of number of steps and pitch on measurement accuracy

We present new extensions of the two-step, triangular-pattern phase-shifting method for different numbers of phase-shifting steps to increase measurement accuracy and to analyze the influence of the number of phase-shifting steps and pitch of the projected triangular intensity-profile pattern on the measurement accuracy. Phase-shifting algorithms to generate the intensity ratio, essential for surface reconstruction, were developed for each measurement method. Experiments determined that higher measurement accuracy can be obtained with a greater number of phase-shifting steps and a lower value of pitch, as long as the pitch is appropriately selected to be divisible by the number of phase-shifting steps and not below an optimal value, where intensity-ratio unwrapping failure would occur.

Dynamic gait stability index based on plantar pressures and fuzzy logic

Stability during locomotion, or dynamic stability, is critical to ensure safe locomotion and a high quality of life. A dynamic stability measure should be easily applied in a clinical setting and must provide a quantitative index that can be used for comparisons over a range of tasks and environments. Plantar foot pressure data acquired by shoe-insole sensors have potential to provide such a measure. To generate a quantitative dynamic gait stability index, six gait parameters were extracted from a commercial plantar pressure measurement system (F-Scan): anterior-posterior (A/P) center of force (CoF) motion, medial-lateral (M/L) CoF motion, maximum lateral position, cell triggering, stride time (ST), and double support time (DST). A fuzzy logic controller combined these six parameters and generated the index. To validate the stability index, 15 healthy subjects performed four tasks intended to induce increasing levels of instability. Fifty-seven gait parameter combinations were assessed to determine the most effective index. A combination of A/P motion, M/L motion, maximum lateral position, and cell triggering parameters was the most consistently effective index across all subjects. However, small changes in ST and DST for able-bodied subjects may have reduced the effectiveness of these measures in the index calculation. The index combining all six parameters should be investigated further with populations with disabilities or pathological gait.

Indicators of dynamic stability in transtibial prosthesis users

An improved understanding of factors related to dynamic stability in lower-limb prosthesis users is important, given the high occurrence of falls in this population. Current methods of assessing stability are unable to adequately characterize dynamic stability over a variety of walking conditions. F-Scan Mobile has been used to collect plantar pressure data and six extracted parameters were useful measures of dynamic stability. The aim of this study was to investigate dynamic stability in individuals with unilateral transtibial amputation based on these six parameters. Twenty community ambulators with a unilateral transtibial amputation walked over level ground, uneven ground, stairs, and a ramp while plantar pressure data were collected. For each limb (intact and prosthetic) and condition, six stability parameters related to plantar center-of-pressure perturbations and gait temporal parameters, were computed from the plantar pressure data. Parameter values were compared between limbs, walking condition, and groups (unilateral transtibial prosthesis users and able-bodied subjects). Differences in parameters were found between limbs and conditions, and between prosthesis users and able-bodied individuals. Further research could investigate optimizing parameter calculations for unilateral transtibial prosthesis users and define relationships between potential for falls and the dynamic stability measures.

Design and Evaluation of a Stance-Control Knee-Ankle-Foot Orthosis Knee Joint

Conventional knee-ankle-foot orthoses (KAFOs) are prescribed for people with knee-extensor muscle weakness. However, the orthoses lock the knee in full extension and, therefore, do not permit a natural gait pattern. A new electromechanical stance-control knee-ankle-foot orthosis (SCKAFO) knee joint that employs a novel friction-based belt-clamping mechanism was designed to enable a more natural gait. The SCKAFO knee joint allows free knee motion during swing and other non-weight-bearing activities and inhibits knee flexion while allowing knee extension during weight bearing. A prototype SCKAFO knee joint was mechanically tested to determine the moment at failure, loading behavior, and wear resistance. The mean maximum resisting moment of the SCKAFO knee joint over five loading trials was 69 Nm plusmn4.9 Nm. The SCKAFO knee-joint strength and performance were sufficient to allow testing on a 90 kg subject at normal walking cadence. Proper function of the new electromechanical knee joint was verified in walking trials of an able-bodied subject

Plantar Pressure Parameters for Dynamic Gait Stability Analysis

Dynamic stability measurement is necessary to evaluate human performance over a variety of locomotor environments. In this paper, the suitability of parameters extracted from plantar-pressure measurements as input into a dynamic stability model was investigated. FScan in-shoe pressure data were collected from 15 subjects as they completed four successively more unstable walking tasks. Six parameters met the criteria of being reliably calculated from plantar pressure data, increasing as the task became more unstable, and relating to past measures from the literature: anterior/posterior centre of force (CoF) position, medio-lateral CoF position, double support time, stance time, cell triggering frequency, and maximum lateral CoF position. These parameters could be combined to create an index of dynamic gait stability

Robot-Manipulator Teleoperation by Markerless Vision-Based Hand-Arm Tracking

This article presents a method of real-time robot-manipulator teleoperation using markerless image-based hand-arm tracking. The markerless tracking is carried out by processing images from two calibrated cameras to estimate in three dimensions (3-D), the positions of the wrist joint, elbow joint, index finger, and thumb. The hand pose (position and orientation) is used to specify the pose of the end-effector of a robot-manipulator in real-time teleoperation. The method permits a natural means of communicating an entire task to a robot, rather than using limited motion commands as with gesture-based approaches, and it has been demonstrated for pick-and-place tasks.