Narayan Srinivasa

Spindle thermal drift measurement using the laser ball bar

Thermally induced errors are major contributors to the overall accuracy of machine tools. An important component of thermally induced errors is the error associated with spindle thermal drifts. In this paper, a novel method is developed to measure spindle thermal drifts in machine tools using a laser ball bar (LBB) as the calibration instrument. The method is implemented on a two-axis CNC turning center. The LBB is used to measure the coordinates of the spindle center and the direction cosines of the spindle axis at various thermal states. The axial, radial, and tilt thermal drifts of the spindle are then computed from the changes in these coordinates. The new method is verified by comparing the spindle drifts measured with the LBB to those measured by capacitance gauges. The results obtained by the new method show good agreement with the capacitance gauge technique. The primary advantage of the new method is the ability to measure the spatial coordinates of the spindle center and direction cosines of the spindle axis with the same instrument used for measurement of the geometric errors of the machine axes.

Grasping and In-Hand Manipulation: Geometry and Algorithms

Abstract. This paper addresses the problem of grasping and manipulating three-dimensional objects with a reconfigurable gripper that consists of two parallel plates whose distance can be adjusted by a computer-controlled actuator. The bottom plate is a bare plane, and the top plate carries a rectangular grid of actuated pins that can translate in discrete increments under computer control. We propose to use this gripper to immobilize objects through frictionless contacts with three of the pins and the bottom plate, and to manipulate an object within a grasp by planning the sequence of pin configurations that will bring this object to a desired position and orientation. A detailed analysis of the problem geometry in configuration space is used to devise simple and efficient algorithms for grasp and manipulation planning. The proposed approach has been implemented and preliminary simulation experiments are discussed.

Automated measurement and compensation of thermally induced error maps in machine tools

In this paper, a direct method of machine tool calibration is adopted to model and predict thermally induced errors in machine tools. This method uses a laser ball bar (LBB) as the calibration instrument and is implemented on a two-axis computerized numerical control turning center (CNC). Rather than individually measuring the parametric errors to build the error model of the machine, the total positioning errors at the cutting tool and spindle thermal drifts are rapidly measured using the LBB within the same experimental setup. Unlike conventional approaches, the spindle thermal drifts are derived from the true spindle position and orientation measured by the LBB. A neural network is used to build a machine model in an incremental fashion by correlating the measured errors with temperature gradients of the various heat sources during a regular thermal duty cycle. The machine model developed by the neural network is further tested using random thermal duty cycles. The performance of the system is also evaluated through cutting tests under various thermal conditions. A substantial improvement in the overall accuracy was obtained.

Grasping and in-hand manipulation : experiments with a reconfigurable gripper

This paper addresses the problem of grasping and manipulating three-dimensional objects with a reconfigurable gripper equipped with two parallel plates whose distance can be adjusted by a computer-controlled actuator. The bottom plate is a bare plane and the top one carries a rectangular grid of actuated pins that can translate in discrete increments under computer control. We propose to use this gripper to immobilize objects through frictionless contacts with three of the pins and the bottom plate, and to manipulate an object within a grasp by planning the sequence of pin configurations that will bring this object to a desired position and orientation. A detailed analysis of the problem geometry in configuration space was used in a previous paper to devise simple and efficient algorithms for grasp and manipulation planning. We have constructed a prototype of the gripper and this paper presents our experiments.

Learning and generalization of noisy mappings using a modified PROBART neural network

Incremental function approximation using the PROBART neural network offers many advantages over conventional feedforward networks. These include dynamic node allocation based on the complexity of the function approximation task, guaranteed convergence, and the ability to handle noise in the training data. However, the PROBART network does not generalize very well to untrained data. In this paper, a modified PROBART is proposed to overcome this deficiency. This modification replaces the winner-take-all mode of prediction of the PROBART with a distributed mode of prediction. This distributed mode enables several neurons to cooperate during prediction and, thus, provides better generalization capabilities even in noisy conditions. Computer simulations are conducted to evaluate the performance of the modified PROBART neural network using three benchmark nonlinear function approximation tasks. The prediction accuracy of the modified PROBART network compares favorably to the PROBART, fuzzy ARTMAP, and ART-EMAP networks for all these tasks.

SOIM: a self-organizing invertible map with applications in active vision

We propose a novel neural network, called the self-organized invertible map (SOIM), that is capable of learning many-to-one functionals mappings in a self-organized and online fashion. The design and performance of the SOIM are highlighted by learning a many-to-one functional mapping that exists in active vision for spatial representation of three-dimensional point targets. The learned spatial representation is invariant to changing camera configurations. The SOIM also possesses an invertible property that can be exploited for active vision. An efficient and experimentally feasible method was devised for learning this representation on a real active vision system. The proof of convergence during learning as well as conditions for invariance of the learned spatial representation are derived and then experimentally verified using the active vision system. We also demonstrate various active vision applications that benefit from the properties of the mapping learned by SOIM.

Learning Multiscale Image Models of 2D Object Classes

This paper is concerned with learning the canonical gray scale structure of the images of a class of objects. Structure is defined in terms of the geometry and layout of salient image regions that characterize the given views of the objects. The use of such structure based learning of object appearence is motivated by the relative stability of image structure over intensity values. A multiscale segmentation tree description is automatically extracted for all sample images which are then matched to construct a single canonical representative which serves as the model of the class. Different images are selected as prototypes, and each prototype tree is refined to best match the rest of the class. The model tree for the class is that tree which is best supported over all the initializations with different prototypes. Matching is formulated as a problem of finding the best mapping from regions of example images to those of the model tree, and implemented as a problem in incremental refinement of the model tree using a learning approach. Experiments are reported on a face image database. The results demonstrate that a reasonable model of facial geometry and topology is learnt which includes prominent facial features.

Efficient learning of VAM-based representation of 3D targets and its active vision applications

There has been a considerable interest in using active vision for various applications. This interest is primarily because active vision can enhance machine vision capabilities by dynamically changing the camera parameters based on the content of the scene. An important issue in active vision is that of representing 3D targets in a manner that is invariant to changing camera configurations. This paper addresses this representation issue for a robotic active vision system. An efficient Vector Associative Map (VAM)-based learning scheme is proposed to learn a joint-based representation. Computer simulations and experiments are first performed to evaluate the effectiveness of this scheme using the University of Illinois Active Vision System (UIAVS). The invariance property of the learned representation is then exploited to develop several robotic applications. These include, detecting moving targets, saccade control, planning saccade sequences and controlling a robot manipulator.

A framework for robot control with active vision using a neural network based spatial representation

Robots that use an active camera system for visual feedback can achieve greater flexibility, including the ability to operate in a dynamically changing environment. Incorporating active vision into a robot control loop involves some inherent difficulties, including calibration, and the need for redefining the goal as the camera configuration changes. In this paper, we propose a novel self-organizing neural network (SOIM) that learns a calibration-free spatial representation of 3D point targets in a manner that is invariant to changing camera configurations. This representation is used to develop a new framework for robot control with active vision. The salient feature of this framework is that it decouples active camera control from robot control. The feasibility of this approach is explored with the help of computer simulations and experiments with the University of Illinois Active Vision System (UIAVS).

A Learning Approach to Fixating on 3D Targets with Active Cameras

Fixation of an active camera pair on a given target requires that the pan and tilt angles of the cameras must be set to bring the target to image centers. However, the calibration needed to achieve a specific configuration of real cameras involves tedious estimation of a number of imaging parameters. Fortunately, this excercise is not essential for fixation if images are acquired and used as feedback during the fixation process to continuously direct the cameras to the target. This paper defines a direct mapping from the changes in the direction of target motion in the image plane to changes in camera angles necessary to reduce the disparity between image center and the image plane target location. The mapping captures camera calibration, as well as other effects such as deviations from the assumed imaging model which are difficult to characterize and capture in calibration. The mapping is formulated as a task in nonlinear function approximation and learnt from real data. For computational efficiency, learning is done at multiple resolutions and using a PROBART network. Experimental results are presented using an active vision system.