To Örebro University

Nowadays, robots have more and more sensors and the technologies allow

using them with less contraints as before. Sensors are important to learn about

the environment. But the sensors can be used for classical conditioning, and

create behavior for the robot. One of the behavior developed in this thesis is

avoiding and predicting obstacles.

The goal of this thesis is to propose a model which consists of developing a

specific behavior to avoid noxious event, obstacles.

This thesis describes the work to make a portable ubiquitous device into a component

in a networked robotic environment. The device can both control and

be controlled in a robotic network as well as used to survey a robotic network.

This is done by building upon existing software for making networked robotic

components and use it on devices with the Android operating system. The software

is the

PEIS-kernel (Physically Embedded Intelligent System) middleware.

The middleware makes it possible to build robotic components called PEIScomponents

that can communicate, work together and share their resources

and capabilities. The result of this work is three-fold,

first a way to make devices

with the Android operating system into PEIS-devices so that they can be

used to control or be controlled as well as survey a PEIS-network.With this the

Android device functionality can be used to extend the capabilities of an entire

PEIS-network.

Secondly, devices with Android get access to a simple and powerful

way to exchange information with any device that is a PEIS-device. The

third

contribution is the capability to make humans a much more integral part

of a PEIS-network as devices with Android are in general easy to use and familiar

to humans. These devices will serve as a very good human - PEIS-network

interaction tool.

Communication blackout is one of the most serious pitfalls of Wireless Sensor Networks (WSN) in industrial automation context. The industrial radio channel exhibits pronounced effects of multipath fading and wireless LAN (WLAN) interference that can potentially lead to temporary communication failures, as well as complete isolation of network devices. The current IWSN standards adopt known countermeasures to cope with the harshness of the radio channel, but they lack solutions specifically oriented to detect blackouts and self-recover the communication fulfilling hard deadline constraints. In this work we focus on the problem of blackout detection with specific interest for the WirelessHART standard, introducing a Blackout Detection Service (BDS) expressly addressed to multi-hop periodic communication with sensors and actuators. The BDS monitors end-to-end acknowledgement messages and builds specific metrics to promptly identify communication outages, enabling three criticality classes. The algorithm is tested in the ns-2 network simulator and results show that the proposed system is able to detect blackout events with reaction delays of the order of 4-5 times the refresh rate of nodes and to discriminate between small and temporary network issues and serious blackout scenarios, opening the field for recovery strategies.

In this thesis work, a novel operating principle for a temperature modulatedelectronic nose is introduced. The main goal is to perform gas discrimination with metal oxide gas sensors in natural, uncontrolled environments where thesensors are exposed to patches of gas only for short periods of time. The proposed Parallelized Temperature Modulated electronic nose (PTM e-nose) allows to speed-up discrimination of gases by measuring in parallel theresponse of n gas sensors of the same type but with a phase-shifted temperaturemodulation cycle. The basic idea is to replicate the base sensor n times with each sensor instance measuring one different nth of the modulation cycle. In this waythe response to the full modulation cycle for one sensor can be recovered from n different sensors in one nth of the time while the chemical response of theindividual sensors is not compromised by a too fast temperature change. The PTM e-nose operating principle is evaluated with an array of four commerciallyavailable tin oxide gas sensors, which are modulated with sinusoids of the same amplitude but phase-shifted by 90 degrees. By addressing gas discrimination in the early stages of the transient response and in the steady state, it is demonstrated that the information contained in one entire modulation cycle can be sufficiently recovered from the responses of the individual sensors.

Patients with Parkinson's disease (PD) need to be frequently monitored in order to assess their individual symptoms and treatment-related complications. Advances in technology have introduced telemedicine for patients in remote locations. However, data produced in such settings lack much information and are not easy to analyze or interpret compared to traditional, direct contact between the patient and clinician. Therefore, there is a need to present the data using visualization techniques in order to communicate in an understandable and objective manner to the clinician. This paper presents interaction and visualization approaches used to aid clinicians in the analysis of repeated measures of spirography of PD patients gathered by means of a telemetry touch screen device. The proposed approach enables clinicians to observe fine motor impairments and identify motor fluctuations of their patients while they perform the tests from their homes using the telemetry device.

Comprehensive analysis of time series data requires a great amount of manual labour, and is not feasible over larger datasets. This thesis concerns the development of a machine learning (ML) solution to automate the detection of potential errors in sequential financial market data for a Swedish government agency, which is a crucial step for reducing the manual workload and maintaining data integrity. Due to the data being imbalanced, with a relatively low occurrence of errors compared to correctly reported values, the primary goal is to aid in the manual data review process by marking reports that are likely to be falsely reported, rather than developing a standalone solution. Two approaches are considered: label-driven error detection and human-based error detection. The first approach targets the classification of actual reported errors through automatically labeled data, and was tackled using various ML models and preprocessing techniques. The final results hinted that the problem of detecting actual errors in the data might not be a viable approach, due to the complexity of said errors and flaws within the labeling process. The second approach aim to mimic the human judgement in detecting potential errors through forecasting. It was evaluated by models such as XGBoost and LSTM-based networks. The results were promising and were deemed as an appropriate solution for the problem at hand.

This paper reports the development and evaluation of a mobile-based telemedicine framework for enabling remote monitoring of Parkinson’s disease (PD) symptoms. The system consists of different measurement devices for remote collection, processing and presentation of symptom data of advanced PD patients. Different numerical analysis techniques were applied on the raw symptom data to extract clinically symptom information which in turn were then used in a machine learning process to be mapped to the standard clinician-based measures. The methods for quantitative and automatic assessment of symptoms were then evaluated for their clinimetric properties such as validity, reliability and sensitivity to change. Results from several studies indicate that the methods had good metrics suggesting that they are appropriate to quantitatively and objectively assess the severity of motor impairments of PD patients.

Spreadsheets, comma separated value files and other tabular data representations are in wide use today. However, writing, maintaining and identifying good formulas for tabular data and spreadsheets can be time-consuming and error-prone. We investigate the automatic learning of constraints (formulas and relations) in raw tabular data in an unsupervised way. We represent common spreadsheet formulas and relations through predicates and expressions whose arguments must satisfy the inherent properties of the constraint. The challenge is to automatically infer the set of constraints present in the data, without labeled examples or user feedback. We propose a two-stage generate and test method where the first stage uses constraint solving techniques to efficiently reduce the number of candidates, based on the predicate signatures. Our approach takes inspiration from inductive logic programming, constraint learning and constraint satisfaction. We show that we are able to accurately discover constraints in spreadsheets from various sources.

The field of mobile robotic telepresence for social communication is in rapid expansion and it is of interest to understand what promotes good interaction. In this paper, we present the results of an experiment where novice users working in health care were given a guided tour while maneuvering a mobile robotic telepresence system for the first time. In a previous study, it was found that subjective presence questionnaires and observations of spatial configurations based on Kendon’s F-formations were useful to evaluate quality of interaction in mobile robotic telepresence. In an effort to find more automated methods to assess the quality of interaction, the study in this paper used the same measures, with the addition of objective sociometric measures. Experimental results show that the quantitative analysis of the sociometric data correlates with a number of parameters gathered via qualitative analysis, e.g. different dimensions of presence and observed problems in maneuvering the robot.

The Internet of Things (IoT) is captivating the society because of its potential to rapidly transform businesses and people’s lives. It is widely believed that IoT will also transform the industrial automation business in terms of improved productivity, less cost, flexibility, and increased revenues. Hence, there are some challenges that needs to be addressed when IoT is introduced to the industrial automation domain. This paperaims to present realistic requirements and highlights identified challenges such as security, interoperability, deterministic and low latency communication, and how the required availability (uptime) can be kept. Moreover, the paper also point out the need of standardization and sustainable business models. The conclusion is that introducing IoT devices and connecting them directly to cloud services is not straightforward for process automation.

There is an increasing interest in the machine learning community to automatically learn feature representations directly from the (unlabeled) data instead of using hand-designed features. The autoencoder is one method that can be used for this purpose. However, for data sets with a high degree of noise, a large amount of the representational capacity in the autoencoder is used to minimize the reconstruction error for these noisy inputs. This paper proposes a method that improves the feature learning process by focusing on the task relevant information in the data. This selective attention is achieved by weighting the reconstruction error and reducing the influence of noisy inputs during the learning process. The proposed model is trained on a number of publicly available image data sets and the test error rate is compared to a standard sparse autoencoder and other methods, such as the denoising autoencoder and contractive autoencoder.

A functional product is an integrated package consisting of hardware, software and a service support system that provides a customer with a certain function and is sold under a performance-based contract that includes a functional availability guarantee. For the availability performance, prediction, optimisation and management of risk are therefore important concerns during product development. This paper describes a software tool that can generate an integrated model of a functional product from its design details and analyse it through simulation to provide availability performance information. The model's application to the analysis of a real industrial system is demonstrated. Such tools are important for the development and widespread adoption of functional products. The resulting analysis gave an indication of a suitable guaranteed functional availability level for the product and could be used to compare the performance of different design options.

Treatment of Parkinson's disease (PD) patients involves major challenges like the large within- and between-patient variability in symptom profiles and the emergence of motor complications. As PD progresses, the symptoms develop slowly and they represent a significant source of disability in advanced patients. During evaluation of treatments and symptoms, both the physician- and patient-oriented outcomes offer complementary information. In addition, quantitative assessments of symptoms using sensing technologies can potentially complement and enhance both patient and clinician perspectives. At Högskolan Dalarna, the Lecturer Mevludin Memedi has developed a telemetry system that assesses symptoms via analysis of self-assessments and motor tests to objectively measure disease-related outcomes and to improve the management of PD.

OBJECTIVES: To develop a method for objective assessment of fine motor timing variability in Parkinson’s disease (PD) patients, using digital spiral data gathered by a touch screen device.

BACKGROUND: A retrospective analysis was conducted on data from 105 subjects including65 patients with advanced PD (group A), 15 intermediate patients experiencing motor fluctuations (group I), 15 early stage patients (group S), and 10 healthy elderly subjects (HE) were examined. The subjects were asked to perform repeated upper limb motor tasks by tracing a pre-drawn Archimedes spiral as shown on the screen of the device. The spiral tracing test was performed using an ergonomic pen stylus, using dominant hand. The test was repeated three times per test occasion and the subjects were instructed to complete it within 10 seconds. Digital spiral data including stylus position (x-ycoordinates) and timestamps (milliseconds) were collected and used in subsequent analysis. The total number of observations with the test battery were as follows: Swedish group (n=10079), Italian I group (n=822), Italian S group (n = 811), and HE (n=299).

METHODS: The raw spiral data were processed with three data processing methods. To quantify motor timing variability during spiral drawing tasks Approximate Entropy (APEN) method was applied on digitized spiral data. APEN is designed to capture the amount of irregularity or complexity in time series. APEN requires determination of two parameters, namely, the window size and similarity measure. In our work and after experimentation, window size was set to 4 and similarity measure to 0.2 (20% of the standard deviation of the time series). The final score obtained by APEN was normalized by total drawing completion time and used in subsequent analysis. The score generated by this method is hence on denoted APEN. In addition, two more methods were applied on digital spiral data and their scores were used in subsequent analysis. The first method was based on Digital Wavelet Transform and Principal Component Analysis and generated a score representing spiral drawing impairment. The score generated by this method is hence on denoted WAV. The second method was based on standard deviation of frequency filtered drawing velocity. The score generated by this method is hence on denoted SDDV. Linear mixed-effects (LME) models were used to evaluate mean differences of the spiral scores of the three methods across the four subject groups. Test-retest reliability of the three scores was assessed after taking mean of the three possible correlations (Spearman’s rank coefficients) between the three test trials. Internal consistency of the methods was assessed by calculating correlations between their scores.

RESULTS: When comparing mean spiral scores between the four subject groups, the APEN scores were different between HE subjects and three patient groups (P=0.626 for S group with 9.9% mean value difference, P=0.089 for I group with 30.2%, and P=0.0019 for A group with 44.1%). However, there were no significant differences in mean scores of the other two methods, except for the WAV between the HE and A groups (P<0.001). WAV and SDDV were highly and significantly correlated to each other with a coefficient of 0.69. However, APEN was not correlated to neither WAV nor SDDV with coefficients of 0.11 and 0.12, respectively. Test-retest reliability coefficients of the three scores were as follows: APEN (0.9), WAV(0.83) and SD-DV (0.55).

CONCLUSIONS: The results show that the digital spiral analysis-based objective APEN measure is able to significantly differentiate the healthy subjects from patients at advanced level. In contrast to the other two methods (WAV and SDDV) that are designed to quantify dyskinesias (over-medications), this method can be useful for characterizing Off symptoms in PD. The APEN was not correlated to none of the other two methods indicating that it measures a different construct of upper limb motor function in PD patients than WAV and SDDV. The APEN also had a better test-retest reliability indicating that it is more stable and consistent over time than WAV and SDDV.

Objective quantification of the upper limb motor timing variability of Parkinson’s disease (PD) patients was evaluated using traces of spirals by groups of patients at different disease stages, stable (S), intermediate (I), advanced (A) and a healthy elderly (HE) group. The approximate entropy (APEN) method of quantifying motor timing variability in time series was applied to capture the amount of irregularity during the spiral drawing process. The APEN score was then normalized by total drawing completion time and used in subsequent analysis. In addition, two previously published methods (WAV and SDDV) were applied on the spiral data. Comparing subject groups’ APEN mean scores, they were found to be significantly different from HE group, for group A (P<0.001) indicating this method’s ability in distinguishing patients at advanced disease stage. Comparing the three methods’ ability to track response to advanced treatment, APEN scores were all significantly different between base-line and levodopa-carbidopa intestinal gel (LCIG) treatment, during the 36 month study period as opposed to WAV and SDDV as they were not significantly improving for all periods. APEN scores were weakly correlated to WAV and SDDV, indicating that they measure different aspects of symptom severity.

Objective: To develop a method for objective quantification of PD motor symptoms related to Off episodes and peak dose dyskinesias, using spiral data gathered by using a touch screen telemetry device. The aim was to objectively characterize predominant motor phenotypes (bradykinesia and dyskinesia), to help in automating the process of visual interpretation of movement anomalies in spirals as rated by movement disorder specialists.

Background: A retrospective analysis was conducted on recordings from 65 patients with advanced idiopathic PD from nine different clinics in Sweden, recruited from January 2006 until August 2010. In addition to the patient group, 10 healthy elderly subjects were recruited. Upper limb movement data were collected using a touch screen telemetry device from home environments of the subjects. Measurements with the device were performed four times per day during week-long test periods. On each test occasion, the subjects were asked to trace pre-drawn Archimedean spirals, using the dominant hand. The pre-drawn spiral was shown on the screen of the device. The spiral test was repeated three times per test occasion and they were instructed to complete it within 10 seconds. The device had a sampling rate of 10Hz and measured both position and time-stamps (in milliseconds) of the pen tip.

Methods: Four independent raters (FB, DH, AJ and DN) used a web interface that animated the spiral drawings and allowed them to observe different kinematic features during the drawing process and to rate task performance. Initially, a number of kinematic features were assessed including ‘impairment’, ‘speed’, ‘irregularity’ and ‘hesitation’ followed by marking the predominant motor phenotype on a 3-category scale: tremor, bradykinesia and/or choreatic dyskinesia. There were only 2 test occasions for which all the four raters either classified them as tremor or could not identify the motor phenotype. Therefore, the two main motor phenotype categories were bradykinesia and dyskinesia. ‘Impairment’ was rated on a scale from 0 (no impairment) to 10 (extremely severe) whereas ‘speed’, ‘irregularity’ and ‘hesitation’ were rated on a scale from 0 (normal) to 4 (extremely severe). The proposed data-driven method consisted of the following steps. Initially, 28 spatiotemporal features were extracted from the time series signals before being presented to a Multilayer Perceptron (MLP) classifier. The features were based on different kinematic quantities of spirals including radius, angle, speed and velocity with the aim of measuring the severity of involuntary symptoms and discriminate between PD-specific (bradykinesia) and/or treatment-induced symptoms (dyskinesia). A Principal Component Analysis was applied on the features to reduce their dimensions where 4 relevant principal components (PCs) were retained and used as inputs to the MLP classifier. Finally, the MLP classifier mapped these components to the corresponding visually assessed motor phenotype scores for automating the process of scoring the bradykinesia and dyskinesia in PD patients whilst they draw spirals using the touch screen device. For motor phenotype (bradykinesia vs. dyskinesia) classification, the stratified 10-fold cross validation technique was employed.

Results: There were good agreements between the four raters when rating the individual kinematic features with intra-class correlation coefficient (ICC) of 0.88 for ‘impairment’, 0.74 for ‘speed’, 0.70 for ‘irregularity’, and moderate agreements when rating ‘hesitation’ with an ICC of 0.49. When assessing the two main motor phenotype categories (bradykinesia or dyskinesia) in animated spirals the agreements between the four raters ranged from fair to moderate. There were good correlations between mean ratings of the four raters on individual kinematic features and computed scores. The MLP classifier classified the motor phenotype that is bradykinesia or dyskinesia with an accuracy of 85% in relation to visual classifications of the four movement disorder specialists. The test-retest reliability of the four PCs across the three spiral test trials was good with Cronbach’s Alpha coefficients of 0.80, 0.82, 0.54 and 0.49, respectively. These results indicate that the computed scores are stable and consistent over time. Significant differences were found between the two groups (patients and healthy elderly subjects) in all the PCs, except for the PC3.

Conclusions: The proposed method automatically assessed the severity of unwanted symptoms and could reasonably well discriminate between PD-specific and/or treatment-induced motor symptoms, in relation to visual assessments of movement disorder specialists. The objective assessments could provide a time-effect summary score that could be useful for improving decision-making during symptom evaluation of individualized treatment when the goal is to maximize functional On time for patients while minimizing their Off episodes and troublesome dyskinesias.

In this thesis, a special class of Recurrent Neural Networks (RNN) is employed

for system identification and predictive control of time dependent systems.

Fundamental architectures and learning algorithms of RNNs are studied

upon which a generalized architecture over a class of state-space represented

networks is proposed and formulated. Levenberg-Marquardt (LM) learning

algorithm is derived for this architecture and a number of enhancements are

introduced. Furthermore, using this recurrent neural network as a system identifier,

a Model Predictive Controller (MPC) is established which solves the optimization

problem using an iterative approach based on the LM algorithm.

Simulation results show that the new architecture accompanied by LM learning

algorithm outperforms some of existing methods. The third approach which

utilizes the proposed method in on-line system identification enhances the identification/

control process even more.

Telepresence robots allow people to explore remote environments, interacting

to other people between long distance. A person only needs to know how to

drive them in order to feel his/herself inside a remote environment. The main

goal of this project is to provide a user interface that improves the user’s depth

perception through a 2D visual feedback and, as a secondary target, to allow

non-expert users to become familiar with the robot control interface. This thesis

shows how a more user friendly interface can reduce the effort needed to drive

the robot properly, especially for non-expert users.

i

Feature selection is the task of selecting a small subset from original features

that can achieve maximum classification accuracy. This subset of features has

some very important benefits like, it reduces computational complexity of learning

algorithms, saves time, improve accuracy and the selected features can be

insightful for the people involved in problem domain. This makes feature selection

as an indispensable task in classification task.

This dissertation presents a two phase approach for feature selection. In the

first phase a filter method is used with “correlation coefficient” and “mutual

information” as statistical measure of similarity. This phase helps in improving

the classification performance by removing redundant and unimportant

features. A wrapper method is used in the second phase with the sequential

forward selection and sequential backward elimination. This phase helps in selecting

relevant feature subset that produce maximum accuracy according to

the underlying classifier. The Support Vector Machine (SVM) classifier (linear

and nonlinear) is used to evaluate the classification accuracy of our approach.

This empirical results of commonly used data sets from the University of

California, Irvine repository and microarray data sets showed that the proposed

method performs better in terms of classification accuracy, number of selected

features, and computational efficiency.

7

In recent years multiprocessor architectures have become mainstream, and multi-core processors are found in products ranging from small portable cell phones to large computer servers. In parallel, research on real-time systems has mainly focused on traditional single-core processors. Hence, in order for real-time systems to fully leverage on the extra capacity offered by new multi-core processors, new design techniques, scheduling approaches, and real-time analysis methods have to be developed.

In the multi-core and multiprocessor domain there are mainly two scheduling approaches, global and partitioned scheduling. Under global scheduling each task can execute on any processor at any time while under partitioned scheduling tasks are statically allocated to processors and migration of tasks among processors is not allowed. Besides simplicity and efficiency of partitioned scheduling protocols, existing scheduling and synchronization methods developed for single-core processor platforms can more easily be extended to partitioned scheduling. This also simplifies migration of existing systems to multi-cores. An important issue related to partitioned scheduling is distribution of tasks among processors which is a bin-packing problem.

In this thesis we propose a partitioning framework for distributing tasks on the processors of multi-core platforms. Depending on the type of performance we desire to achieve, the framework may distribute a task set differently, e.g., in an application in which tasks process huge amounts of data the goal of the framework may be to decrease cache misses.Furthermore, we propose a blocking-aware partitioning heuristic algorithm to distribute tasks onto the processors of a multi-core architecture. The objective of the proposed algorithm is to decrease blocking overhead of tasks which reduces the total utilization and has the potential to reduce the number of required processors.Finally, we have implemented a tool to facilitate evaluation and comparison of different multiprocessor scheduling and synchronization approaches, as well as different partitioning heuristics. We have applied the tool in the evaluation of several partitioning heuristic algorithms, and the tool is flexible to which any new scheduling or synchronization protocol as well as any new partitioning heuristic can easily be added.

Motivated by an analogy with matrix factorization, we introduce the problem of factorizing relational data. In matrix factorization, one is given a matrix and has to factorize it as a product of other matrices. In relational data factorization (ReDF), the task is to factorize a given relation as a conjunctive query over other relations, i.e., as a combination of natural join operations. Given a conjunctive query and the input relation, the problem is to compute the extensions of the relations used in the query. Thus, relational data factorization is a relational analog of matrix factorization; it is also a form inverse querying as one has to compute the relations in the query from the result of the query. The result of relational data factorization is neither necessarily unique nor required to be a lossless decomposition of the original relation. Therefore, constraints can be imposed on the desired factorization and a scoring function is used to determine its quality (often similarity to the original data). Relational data factorization is thus a constraint satisfaction and optimization problem. We show how answer set programming can be used for solving relational data factorization problems.

In this thesis work, we evaluate an object recognition system for an autonomous

wheel loader, to detect objects in its vicinity, in particular an articulated hauler

truck, by using an interest point extraction method that explicitly considers object

borders information, combined with a feature descriptor known as Normal

Aligned Radial Features (NARF) in 3D point cloud data. The object recognition

technique relies on extraction of NARF from range images (computed from

point clouds) for both model(hauler) and the scene. The technique used is robust

feature matching where the extracted model features are mapped on to

the scene containing the model and then seeking for a best transformation that

aligns the model with respect to the scene.

In this context we conducted several experiments with many number of 3D

scans obtained from the laser scanner mounted on the top of an autonomous

wheel loader to analyze the accuracy of the object recognition system. Finally

we demonstrated the results, as the system is able to recognize the hauler from

any view point.

vii

Determining the state of a tricone drill bit in mining operations today is done through manual inspections either, with visual inspection or analyzing the rate of penetration. Manual inspections are time consuming and inefficient, creating unnecessary halts in the drilling operations. In order to create an autonomous mine, with high efficiency and better security, automatic detection of drill bit wear is required. This master’s thesis is done in collaboration with Epiroc, a Swedish company that specializes in producing mining, construction and infrastructure equipment and is a leading innovator in autonomous mining. Epiroc can get an objective estimate of drill bit wear that could assist service technicians. This master’s thesis analyzes different approaches to extract the features of different teeth wear, trains and evaluates different machine learning models, both created from scratch and already trained on a different dataset, and suggests a model proposal that can be investigated in the future. The results of the investigating work show that there already exist methods to classify wear of drill bits but they only detect if a drill bit is new or completely worn. An implemented CNN model showed 78 % accuracy in classifying drill bits by looking at individual teeth and being able to differentiate between usable and unusable drill bits. This method offers a more precise way to determine drill bit condition, showing promising results even though the model is trained on low quality images and demonstrates that autonomous wear detection on drill bits using computer vision has potential.

This thesis presents a study of the signed distance function as a three-dimensional

implicit surface representation and provides a detailed overview of its

different properties. A method for generating such a representation using the

depth-image output from a Kinect camera is reviewed in detail. In order to improve

the quality of the implicit function that can be obtained, registration of

multiple sensor views is proposed and formulated as a camera pose-estimation

problem.

To solve this problem, we first propose to minimize an objective function,

based on the signed distance function itself. We then linearise this objective

and reformulate the pose-estimation problem as a sequence of convex optimization

problems. This allows us to combine multiple depth measurements

into a single distance function and perform tracking using the resulting surface

representation.

Having these components well defined and implemented in a multi-threaded

fashion, we tackle the problem of object detection. This is done by applying the

same pose-estimation procedure to a 3D object template, at several locations,

in an environment reconstructed using the aforementioned surface representation.

We then present results for localization, mapping and object detection.

Experiments on a well-known benchmark indicate that our method for localization

performs very well, and is comparable both in terms of speed and

error to similar algorithms that are widely used today. The quality of our surface

reconstruction is close to the state of the art. Furthermore, we show an

experimental set-up, in which the location of a known object is successfully

determined within an environment, by means of registration.

i

Quadrocopter is an aerial vehicle platform which has become very popular

among researchers in the recent past due to the advantages it offers over conventional

helicopters. Quadrocopter is very simple and unique, but it is inherently

unstable from aerodynamics point of view. In recent past researchers

have proposed many control schemes for quadrocopter. In this thesis we present

fuzzy logic controller for quadrocopter. After brief introduction brief hardware

details are given that is used in this thesis. After that design procedure for the

fuzzy controller is presented. Then the designed fuzzy controller is tested in

Hardware In Loop (HIL) setup. The experimentation to validate the functionality

and applicability of the designed controller were performed in contrained

setup due to some technical problems. The results of the experiments were satisfactory

and it is concluded that it is possible to stabilize quadrocopter with

fuzzy logic controller.

Parkinsons disease (PD) is currently incurable, however the proper treatment can ease the symptoms and significantly improve the quality of patients life. Since PD is a chronic disease, its efficient monitoring and management is very important. The objective of this paper is to investigate the feasibility of using the features and methodology of a spirography device, originally designed to measure early Parkinsons disease (PD) symptoms, for assessing motor symptoms of advanced PD patients suffering from motor fluctuations. More specifically, the aim is to objectively assess motor symptoms related to bradykinesias (slowness of movements occurring as a result of under-medication) and dyskinesias (involuntary movements occurring as a result of over-medication). The work combines spirography data and clinical assessments from a longitudinal clinical study in Sweden with the features and pre-processing methodology of a Slovenian spirography application. The target outcome was to learn to predict the “cause” of upper limb motor dysfunctions as assessed by a clinician who observed animated spirals in a web interface. Using the machine learning methods with feature descriptions from the Slovenian application resulted in 86% classification accuracy and over 90% AUC, demonstrating the usefulness of this approach for objective monitoring of PD patients.

Humanoid robotics is a challenging and promising research field. Legged locomotion

is one of the most important aspects of it. In spite of the progress

achieved in the last years in control of walking robots, many problems are yet

to be resolved. The inherent complexity of such robots makes their control a

difficult task even on the modern hardware. In order to address this issue approximate

models and high performance algorithms are employed. This thesis

is focused on the model predictive control of a walking bipedal robot, which

is approximated by an inverted pendulum, using online optimization. A special

emphasis is made on the solvers that exploit the structure of quadratic optimization

problems in the context of model predictive control. Two methods for

solution of these problems are implemented: primal active set and primal logarithmic

barrier methods. They are tested and compared in a simulation and on

a humanoid robot. A software module for control of the Nao humanoid robot

is developed for this purpose.

The highly percipient nature of human mind in avoiding sensory overload is a crucial factor which gives human vision an advantage over machine vision, the latter has otherwise powerful computational resources at its disposal given today’s technology. This stresses the need to focus on methods which extract a concise representation of the environment inorder to approach a complex problem such as visual mapping. This article is an attempt of creating a mapping system, which proposes an architecture that combines task-specific and saliency driven approaches. The proposed method is implemented on a warehouse robot. The proposed solution provide a priority framework which enables an industrial robot to build a concise visual representation of the environment. The method is evaluated on data collected by a RGBD sensor mounted on a fork-lift robot and shows promise for addressing visual mapping problems in industrial environments.

Objective: The aim of this study is to develop and evaluate methods for quantifying motor symptoms in Parkinson’s disease (PD) using combined upper limb motor test data collected during tapping and spiral drawing tasks by a smart phone.

Background: PD is a multidimensional and complex disorder affecting motor and non motor functionalities. Assessments of PD symptoms are usually done by clinical rating scales. One of them is the Unified PD Rating Scale (UPDRS) developed to provide comprehensive, efficient and flexible means to monitor PD-related disability and impairments. It has been the most commonly used rating scale. It is composed of four main parts where the third part is designed for rating of motor symptoms. However, UPDRS has relatively poor inter-rater reliability. Another scale that is used to grade motor function of patients is Treatment Response Scale (TRS). A limitation is that there is no general agreement on which parts of the symptomatology should be included in the TRS score. However, the scales have low intra- and inter-rater reliability and their use is limited, as they are only used during in-clinic observations.

Methods: Participants: Nine-teen patients diagnosed with PD and 22 healthy controls were recruited in a single center, open label, single dose clinical observational study in Sweden. Simultaneous clinical- and smartphone-based measures were collected up to 15 times following a single levodopa/carbidopa morning dose (50% over normal to induce dyskinesias).

Clinical assessment: Subjects were asked to perform standardized motor tests in accordance with UPDRS and were videotaped. The videos were blindly rated by three movement disorder specialists. The ratings were given based on TRS ranging from -3, 'Very Off' to 0, 'On' to +3, 'Very dyskinetic', three UPDRS motor items (item 23, finger taps; item 25, rapid alternating movements of hands, item 31, body bradykinesia and hypokinesia) and dyskinesia score. Means of the three specialists' assessments per time point on these scales were used in subsequent analysis.

Smartphone-based data collection: On each test occasion, the subjects performed upper limb motor tests (tapping and spiral drawings) using a smartphone. The subjects were instructed to perform the tests using an ergonomic pen stylus with the device placed on a table and to be seated in a chair. During tapping tests, the subjects were asked to alternately tap two fields (as shown in the screen of the device) as fast and accurate as possible, using first right hand and then left hand. Each tapping test lasted for 20 seconds. During spiral tests, the subjects were instructed to trace a pre-drawn Archimedes spiral as fast and accurate as possible, using the dominant hand. The spiral test was repeated three times per test occasion. The smartphone recorded both position and time-stamps (in milliseconds) of the pen tip.

Data processing and analysis: The raw tapping and spiral data were processed with time series analysis methods, including both time- and frequency-domains methods. Nineteen and 22 spatiotemporal features were extracted from spiral and tapping data, respectively. Features were calculated to represent various kinematic quantities during the motor tests such as acceleration, speed, time delay, and distance. The features from both tapping and spiral data were used in a Principal Component Analysis and 7 principal components (PCs) were retained, which in turn were used as inputs to a Support Vector Machines (SVM) to be mapped to mean clinical ratings. The analyses were performed with a stratified 10-fold cross-validation. Test-retest reliability of the spiral tests were assessed after calculating correlations between the first PCs for the three spiral tests and then calculating the mean of all possible correlations.

Results: The correlation coefficients between SVM predictions and mean clinical ratings were as follows: 0.59 for TRS, 0.6 for dyskinesia score, 0.52 for item 23 of UPDRS (finger taps), 0.47 for item 25 of UPDRS (rapid alternating movements of hands), and 0.57 for item 31 of UPDRS (body Bradykinesia and Hypokinesia). The spiral test had a good test-retest reliability with a coefficient of 0.84, indicating that spiral scores are stable and consistent over time. When assessing the ability of the PCs to distinguish between patients and healthy controls the means of 3 out of 7 PCs (PC1, PC2 and PC4) were different between the two groups (p<0.005).

Conclusions: The upper limb motor tests of the smartphone were able to capture important and relevant symptom information of the clinical rating scales. The methods for quantifying the upper limb motor symptoms of PD patients: had adequate correlations to clinical ratings were able to differentiate between movements of patients and healthy controls, and(Spiral tests) had good test-retest reliability.

To apply industrial robots in friction stir welding (FSW) for difficult-to-weld materials and alloys has until recently been a proposed task. However, yet the laboratory experiments did not provide a feasible industrial application. We describe our approach to modify and provide an industrial robot with FS-welding capacity by modifying a standard industrial robot through replacing its sixth axis with FSW related equipment. The emphasis is on achieving reasonable welding speed and path complexity in 3D space. As significant force is needed for FSW and at the same time position precision has to be kept, the control problems become complicated. We demonstrate our first experiments, highlighting this problem and point some possible solutions

In certain robot applications, path planning has to be viewed, not only from a motion perspective, but also from a process perspective. In 3-dimensional Friction Stir Welding (FSW) a properly planned path is essential for the outcome of the process, even though different control loops compensate for various deviations. One such example is how sharp path intersection is handled, which is the emphasis in this paper. We propose a strategy based on Hermite and Bezier curves, by which G1 continuity is obtained. The blending operation includes an optimization strategy in order to avoid high second order derivatives of the blending polynomials, yet still to cover as much as possible of the original path.

In certain robot applications, path planning has to be viewed, not only from a motion perspective, but also from a process perspective. In 3-dimensional Friction Stir Welding (FSW) a properly planned path is essential for the outcome of the process, even though different control loops compensate for various deviations. One such example is how sharp path intersection is handled, which is the emphasis in this paper. We propose a strategy based on Hermite and Bezier curves, by which G1 continuity is obtained. The blending operation includes an optimization strategy in order to avoid high second order derivatives of the blending polynomials, yet still to cover as much as possible of the original path.

The growing trend towards adoption of flexible and heterogeneous, parallel computing architectures has increased the challenges faced by the programming community. We propose a method to program an emerging class of reconfigurable processor arrays by using the CSP based programming model of occam-pi. The paper describes the extension of an existing compiler platform to target such architectures. To evaluate the performance of the generated code, we present three implementations of the DCT algorithm. It is concluded that CSP appears to be a suitable computation model for programming a wide variety of reconfigurable architectures.

This paper presents a modified version of the Lucas-Kanade (LK) method that incorporates depth maps to enhance the accuracy of tracking. Traditional LK methods, widely adopted for optical flow estimation, typically rely on intensity values of pixels to estimate motion. However, such approaches are often vulnerable to challenges presented by texture-less regions, perspective shifts, and occlusions. By integrating depth maps into the LK framework, this method adds a valuable dimension of spatial information that alleviates these limitations. This approach utilizes depth values to filter out layers of the scene not relevant to the point of interest, providing a more robust descriptor for tracking applications. Experimental results on custom datasets demonstrate that the depth aware LK method generally outperforms conventional LK algorithms in terms of accuracy and robustness. Furthermore, potential applications and the broader implications of said application are discussed.

This project, completed in collaboration with Piktiv AB, aimed to develop an automated surface hole-filling feature for static meshes in Unreal Engine 4.27, with the goal of making repaired surfaces visually indistinguishable from their surrounding areas. The solution was primarily designed to address holes that arose from, but were not limited to, the use of Reduction Settings within Unreal Engine on static meshes. The functionality encompassed four key stages: boundary detection, where all holes on the mesh were identified; triangulation, which involved patching the hole using vertices from the boundary; refinement, entailing the addition of vertices and triangles to the patched area to mimic the density of the surrounding surface; and fairing, which smoothed the patched surface. Additionally, the project introduced a straightforward method for determining the texture coordinates of newly added vertices and a technique for ensuring that triangle normals correctly faced outward from the mesh. The Static Mesh Hole Filler, as implemented, demonstrates efficiency in filling an arbitrary amount of small, planar holes, which commonly result from polygon reduction using Reduction Settings in Unreal Engine. However, this function falls short in preserving unique texture details and maintaining the curvature of surfaces when dealing with larger holes. This limitation necessitates users to seek alternative methods for effectively repairing the mesh.

Recently we proposed occam-pi as a high-levellanguage for programming coarse grained reconfigurable architectures. The constructs of occam-pi combine ideas from CSPand pi-calculus to facilitate expressing parallelism, communication, and reconfigurability. The feasability of this approachwas illustrated by developing a compiler framework to compile occam-pi implementations to the Ambric architecture.

In this paper, we demonstrate the applicability of occam-pif or programing an array of functional units, eXtreme ProcessingPlatform (XPP). This is made possible by extending the compilerframework to target the XPP architecture, including automatic floating to fixed-point conversion. Different implementations of a FIR filter and a DCT algorithm were developed and evaluated on the basis of performance and resource consumption. The reported results reveal that the approach of using occam-pito program the category of coarse grained reconfigurable architectures appears to be promising. The resulting implementations are generally much superior to those programmed in C and comparable to those hand-coded in the low-level native language NML.

Wireless communication has gained more interest in industrial automation due to flexibility, mobility, and cost reduction. Wireless systems, in general, require additional and different engineering and maintenance tasks, for example cryptographic key management. This is an important aspect that needs to be addressed before wireless systems can be deployed and maintained efficiently in the industry. In this paper, we take an holistic approach that addresses safety and security regardless of the underlying media. In our proposed framework we introduce security modules which can be retrofitted to provide end-to-end integrity and authentication measures by utilizing the black channel concept. With the proposed approach, we can extend and provide end-to-end security as well as functional safety using existing automation equipment and standards, such as Profisafe, Profinet IO, and WirelessHART. Furthermore, we improve the WirelessHART standard with periodic and deterministic downlink transmissions to enable efficient usage of wireless actuators, as well as improving the performance of functional safety protocols.