Tapio Heikkilä

Flexible workobject localization for CAD-based robotics

In this paper a method to locate work objects with splined surfaces and estimate the spatial uncertainties of the estimated parameters is presented. The reference B-spline surface patch is selected from a work object CAD-model and is defined in the form of control vertices. The process includes the hang-eye calibration of the sensor, determination of the work object localization and surface treating, e.g. inspection. The hand-eye calibration and work object localization are carried out using the Bayesian form estimation with sensor fusion. Use of the recursive sensor fusion method makes calibration more flexible and accurate in handling large data sets. The spatial uncertainties in the form of eigenvalues in the direction of the eigenvectors are analyzed from the error covariance matrices of the estimated parameters.

A Holonic Shot-blasting System

Holonic features such as cooperative and autonomous behaviour can ensure the flexibility and robustness of a manufacturing system. When a holonic system is to be implemented, several important issues have to be resolved: How to describe the task to be completed? How to decompose a task for different production devices? How to optimise the workflow and co-ordinate the production?

Robotic deburring system of foundry castings based on flexible workobject localization

This paper presents methods to improve flexibility and accuracy of deburring of castings. We apply several methods including off-line programming of the deburring paths, accurate localization of the work object, surface measuring of the work object and a force-guided motion control during the deburring task. The eye-in-hand calibration as well as the localization of the work object we carry out using the Bayesian-form estimation method with recursive sensor fusion. As a result from the work object localization we obtain a 3 DOF location increment (position difference between the simulation model and actual workcell) and actual deburring paths are corrected using that increment. The simulation phase includes octree-based collision check and the faceting uses octree principle too. The paper includes results from actual tests which are promising. The methods are designed to be easy-to-implement in any industrial robot.

Effect of noise level for estimating the spatial uncertainties of work object localization

In this paper, we present a method to estimate spatial uncertainties of a localized workobject using Bayesian estimation. We approch the problem of a sensor eye-in-hand calibration with error covariances by comparing the covariance propagation with Monte Carlo simulation and actual tests when the system noise level is changing. The spatial uncertainties are analysed using eigenvalues of the covariances in the direction of the respective eigenvectors. Results from the comparison between the different methods gives encouraging results and we believe that covariance propagation can be used in uncertainty estimation in different levels of noise.

Reduction of spatial uncertainties as a criteria for sensing planning of hand-eye calibration

In this paper, we present a method to generate a set of samples which decreases the uncertainties of the estimated parameters. The goal is to carry out hand-eye calibration, i.e. estimate the transformation from wrist of the robot to the coordinate origin of the sensor attached in the wrist of the robot. Using the presented method we decrease the spatial uncertainties and avoid cases where the set of samples are poor and estimator fails or gives an unreliable estimate for both parameters and related uncertainties. This is important especially in the noised conditions and the cases where only a sparse set of samples is available, e.g. hand-eye calibration with a singlepoint laser rangefinder tactile sensor. The planning method is compared with pattern and random sets of samples and results for the new method are promising.

Sensing Planning of Calibration Measurements for Intelligent Robots

1.1 Overview of the problem Improvement of the accuracy and performance of robot systems implies both external sensors and intelligence in the robot controller. Sensors enable a robot to observe its environment and, using its intelligence, a robot can process the observed data and make decisions and changes to control its movements and other operations. The term intelligent robotics, or sensor-based robotics, is used for an approach of this kind. Such a robot system includes a manipulator (arm), a controller, internal and external sensors and software for controlling the whole system. The principal motions of the robot are controlled using a closed loop control system. For this to be successful, the bandwidth of the internal sensors has to be much greater than that of the actuators of the joints. Usually the external sensors are still much less accurate than the internal sensors of the robot. The types of sensors that the robot uses for observing its environment include vision, laser-range, ultrasonic or touch sensors. The availability, resolution and quality of data varies between different sensors, and it is important when designing a robot system to consider what its requirements will be. The combining of information from several measurements or sensors is called sensor fusion. Industrial robots have high repeatable accuracy but they suffer from high absolute accuracy (Mooring et. al. 1991). To improve absolute accuracy, the kinematic parameters, typically Denavit – Hartenberg (DH) parameters or related variants can be calibrated more efficiently, or the robot can be equipped with external sensors to observe the robot’s environment and provide feedback information to correct robot motions. With improved kinematic calibration the robot’s global absolute accuracy is improved. While using external sensors the local absolute accuracy is brought to the accuracy level of the external sensors. The latter can also be called workcell calibration. For kinematic calibration several methods have been developed to fulfil the requirements of several applications. There are two main approaches for calibration (Gatla et. al. 2007): open loop and closed loop. Open loop methods use special equipment such as coordinate measuring machines or laser sensors to measure position and orientation of the robot end-effector. These methods are relatively expensive and timeconsuming. The best accuracy will be achieved when using these machines as Visual Servoing tools where they guide the end-effector of the robot on-line (Blank et. al. 2007). Closed loop methods use robot joint measurements and end-effector state to form closed loop equations for calculating the calibration parameters. The state of the end effector can be

An Assistive Surgical MRI Compatible Robot - First Prototype with Field Tests

Magnetic Resonance Imaging (MRI) is superior to other imaging modalities in detecting diseases and pathologic tissue in the human body. The excellent soft tissue contrast allows better delineation of the pathologic and surrounding structures. For example, brain surgery requires exact three-dimensional orientation to piece together anatomical and pathological locations inside the brain. The target location can be seen in the preoperative MRI and neuroradiologists can give assessments, e.g., of tumor nature. Still, factors affecting the resection technique e.g. density of neovasculature and consistency of tumor tissue cannot always be evaluated beforehand. Intraoperative MRI (IMRI) complementing preoperative MRI – is continuously being developed to give additional information to the neurosurgeon (Tuominen et al., 2002, Yrjana, 2005).

A Concept for Isles of Automation

1.1 Approach to the topic There has been a major change in the manufacturing after year 2000. At the same time when production is struggling with tougher price competition, the requirements for flexibility are increasing. Life cycles of the products are getting shorter, the variety of products is increasing, and production costs should be decreased at the same time when there is no good technical solution on the market to answer to this. Customer assumes to get a tailored solution with the same price and delivery as a mass product previously. Using the current manufacturing technologies response to the needs of the market is getting increasingly difficult. Even if the modern manufacturing systems include a remarkable amount of ICT technologies, the flexibility of a human worker is very challenging to reach. Although robotic systems are classified as a flexible production technology, in practice the current robotic implementations are concentrated to high volume production (Naumann et. al 2006). Only a few industrial solutions have been installed for short and single series production. The main obstacle in installing robots for short series production is the amount of product-specific costs caused by the manual work still required for using the robot system (Sallinen et. al 2004). In most of the cases each work phase and process for each product has to be programmed, and the function of auxiliary equipment is usually based on part-specific geometry e.g. no exceptions beyond the designed parts are allowed. If the volumes of the product are low, the effective utilization of a robot assumes that it is applied to a large variety of parts, or to a large amount of different work phases, to bring the robot utilization rate to a decent level typical of the SME industry. In addition, parts to be processed in the production environment very often have complicated forms which make manipulation more challenging. This is the case especially in the short series production when most of the manufacturing processes are solved relatively well at the moment.

Planning of collision-free paths for holonic Blastman robots

Collision freeness is an essential feature in off-line programming of robots. It relates to safe and efficient operation, and becomes even more critical, when two or more robots are sharing the same work space, which is typical for so called holonic robot systems. Here we are reporting a simple and fast path planning algorithm based on configuration space method, suitable for holonic robot applications. The basic version of the algorithm can be found from the literature, and we have extended and modified it to improve its feasibility. Simulation and experimental tests have been carried out successfully.