University competence centres (UCCs) are created to enhance university-industry collaboration and knowledge sharing among collaborating partners. This study investigates the organisation of knowledge sharing among firms in UCCs through a qualitative case study of UCCs with or without a focus on research in their activities. Data collection was done through interviews and observations over a period of 24 months. While the findings indicate that both types of UCCs are non-neutral, they also reveal several different characteristics that appear primarily based on a strong tie either to the first (education) or the second (research) mission of academia. Although both types of UCCs act to build a common meaning among participating organisations, the focus on the first or the second mission leads to this meaning is primarily being constructed in the firm-to-firm or university-to-firm interfaces, respectively. Whereas cross-industry knowledge sharing is emphasised by both types of centres, it is thus more strongly emphasised by UCCs without a focus on research as it helps to avoid harmful effects of knowledge spillovers. The focus on the first mission also appears able to sustain the organisation of knowledge ecosystems created by UCCs without a focus on research in a prefigurative form, which is otherwise typically transient. Furthermore, the challenges to sustainability are different, with centres focused on research being pre-occupied with funding issues, while centres not focused on research leveraging on others means to maintain the interest of industry. The findings contribute to innovation management research and practice by refining current understanding of processes and practices of university-industry collaboration, and how they contribute to facilitate (cross-industry) collaboration and knowledge transfer. Given that university-industry collaboration is often promoted in national innovation policies to create value for society as whole, our findings contribute towards enabling organisations, managers as well as governments to take more informed actions when engaging in such collaborations.

This study investigates the role of University–Industry Collaboration (UIC) in enabling cross-industry knowledge sharing, focusing on the implications of structural social capital within these collaborations. Through a multiple case study involving three distinct UIC structures, we explore how the nature of these collaborations, whether research-focused or networking-focused, influences cross-industry knowledge sharing. Our findings reveal that research-focused UIC, characterized by formal and structured interactions, primarily leverages cognitive social capital to abstract and disseminate knowledge across industries. Conversely, networking-focused UIC, which operates through informal and less-structured channels, relies mainly on relational social capital to foster direct, trust-based knowledge sharing between industries. We identify key enablers and distinctions in these UIC structures, illustrating how they shape cross-industry knowledge sharing. Specifically, we propose a conceptual model, highlighting the mediating role of social capital dimensions and the moderating effects of interest and funding. This model offers new insights into the relationship between structural social capital and cross-industry knowledge sharing in UIC. From a managerial perspective, our study suggests that whether firms engaged in UIC can leverage UIC structures for cross-industry knowledge sharing largely comes down to their own strategical development and management of social capital. From a policy perspective, our study suggests that augmenting already existing policy toward cross-regional knowledge sharing by considering aspects of networking-focused UIC and cross-industrial knowledge sharing, could leverage gains from UIC.

A viable option to reduce global warming related to internal combustion engines is to use renewable fuels, for example methanol. However, the risk of knocking combustion limits the achievable efficiency of SI engines. Hence, most high load operation is run at sub-optimal conditions to suppress knock. Normally the fuel is a limiting factor, however when running on high octane fuels such as methanol, other factors also become important. For example, oil droplets entering the combustion chamber have the possibility to locally impact both temperature and chemical composition. This may create spots with reduced octane number, hence making the engine more prone to knock. Previous research has confirmed a connection between oil droplets in the combustion chamber and knock. Furthermore, previous research has confirmed a connection between oil droplets in the combustion chamber and exhaust particle emissions. However, the co-variation between oil originating particle emissions and knock has not been investigated. The current study examines the connection between knock and particle number in the exhaust, when running on fuel with low soot production. A single cylinder spark ignited heavy-duty engine was used. It was equipped with port fuel injection and fueled with methanol, which produces very little soot at lambda 1. Consequently, the measured exhaust particle numbers were assumed to origin essentially from engine oil. Three grades of oil, in combination with three piston ring configurations, were used to vary the amount of oil entering the combustion chamber. Results from knock limited operation at both medium and high engine load showed that an increased number of particles in the exhaust was associated with an increased likelihood of knock. The authors find the hypothesis that an increase in particle number correlates with an increase in auto-ignition tendency to be confirmed.

Mecanum wheels were invented in the year 1972 and today the wheels are commonly used in industries where you need to orient machinery in small spaces. Containerloaders are an example which use several powerful mecanum wheels for rotation and placing of containers. The wheels are also used in hobby purposes because of their flexible and odd movement capabilities. Cannons, or turrets, are used for several different tasks, for example in the weapons industry, but also in toys and such. This paper will discuss the regulation of the mecanum wheels as well as the construction of a RC-car in combination with a pneumatic turret on its roof, entirely motivated by pure interest and for the fun of it. The paper also processes the choice of components, different construction solutions and lastly it presents the final construction that also fulfills the set targets and answers the stated research questions. A PI controller was chosen for the wheel regulation, which includes a proportional part to quickly return to a reference value and an integral part to eliminate steadystate errors.

This article presents a novel emotionally expressive robot platform targeting social engagement with children. This platform was implemented in accordance with UNICEF's policy guidance on artificial intelligence (AI) for children, focusing on factors such as safety, transparency, reliability and explainability. The robot prototype is presented from a design and development perspective, outlining all utilized electromechanical components that enable its 11 degrees-of-freedom and sensing functions. Preliminary evaluation results are provided in terms of dependability and expressiveness of basic emotions, thus demonstrating the robot's potential to facilitate trustworthy and secure interactions with children.

This paper focuses on the application of soft robotics in the area of social interaction, and presents a modular approach on the design of a soft robotic neck for integration with the social robot HARU. The proposed design incorporates soft robotics and additive manufacturing principles, enhancing safety through compliance for absorbing contacts with users or objects, while modularity allows for easy replacement or upgrade to meet HARU's specific application requirements. The paper discusses the conceptual design specifics of the soft robotic neck and provides an overview of the prototype development stages and its main functionalities.

This bachelor thesis is aimed at creating an improved version of traditional two-story bicycle racks, common in bicycle rooms. The purpose is to remove the need for heavy lifting when loading and unloading bicycles, while simultaneously enhancing safety and utilization. The solution presented in this project is to create an automated rotary system that is connected to an identification system, pairing ID tags to the specific bike slots. The simplistic idea is that one scans their tag, and a motor rotates the whole bike rack to present one’s bike slot on ground level for easy access. The system also aims to increase security by keeping the storage unit behind a hatch that only opens when a tag is scanned. The research involved analyzing the technical research landscape using sources from Google Scholar and company websites, as well as prior bachelor theses. The construction process included CAD modelling, 3D-printing of components, and the implementation of mechanical bicycle parts. The finished prototype has a 3D-printed chassis with assembled motors, sensors and mechanical components, coupled with a microprocessor that controls the system with the help of Arduino C++ code. It provides a deviation of ≤3,54 mm per revolution, and an average speed of retrieval of 1,69 s from the moment of scanning any key tag. The prototype achieves great functionality and the concept has, with appropriate adaptations, potential for real-life application.

This project involved constructing a prototype of a drink dispenser that allows for the ordering and serving of a number of popular mixed drinks. The flow from the dispenser is controlled by solenoid valves, which in turn are controlled by a microcontroller. The user orders through a menu on a touchscreen that sends signals to the microcontroller. Upon ordering, a reaction test is activated to verify that the user is sober enough to place an order. This test is inspired by the reaction test of the electric scooter-company Voi Technology and involves the user clicking on symbols that appear on the screen for a short period of time. The prototype should also be able to determine if a bottle needs to be replaced and notify the staff, this is achieved by a counter in the program code tracking how many centiliters are poured from the bottle and opening a warning-window on the screen after a certain amount. Due to manufacturing defects in four out of six purchased solenoid valves, the prototype was reduced to only serve one cocktail. To still offer choices when ordering via the touchscreen, the user is instead allowed to choose the amount of alcohol and mixer for their drink. To test the precision capability of the prototype, the deviation in dispensing 4 centiliters of liquid was examined. These tests showed that the outlet with the lowest accuracy had an average deviation of 6.25 percent.

Systems of Systems (SoSs) integrate many critical systems our society relies on. In designing individual systems, stakeholders use bespoke protocols, custom information models, and proprietary components with limited computational resources from various vendors. We present a reference architecture that allows multiple stakeholders to carry out a flexible integration without giving up control to a single entity in the presence of the aforementioned limitations. Our architecture relies on rule engines and graph data model to integrated systems flexibly even when black-box components are used. At the same time, a federation of the rule engines allows each stakeholder to retain control over the rules that reflect their policies. We also rely on a common information model based on ontologies to account for the information model mismatch and reduce the duplication of integration efforts. Moving rule execution to the standalone rule engines allows deployment in resource-constrained and proprietary environments. A uniform application programming interface (API) is used to integrate rule engines across systems as well as components within each system with a respective rule engine. We also present a novel algorithm to determine dependencies across rules deployed in different rule engines within the federation. This allows domain experts to develop rules as usual without having to deal with the distributed aspect of the system. We also present a proof of the sufficient condition to ensure all necessary notifications will be sent to ensure correct rule activation across different rule engines. Compared to other systems involving distributed rules, the proposed architecture is well-suited for the integration of transactional workloads commonly found in enterprises. The qualitative evaluation based on the Architecture Tradeoff Analysis Method (ATAM), applied to a telecommunications use case, shows that the architecture possesses the “interoperability”, “modifiability”, and the “functional completeness” quality attributes with a trade-off around rule expressiveness. The quantitative evaluation demonstrates speedup over the single-node setup in most scenarios except in case of highly optimized rules and a poor network performance simultaneously (tr<10ms, tn=100ms).

High accuracy autonomous positioning systems are becoming increasingly important in today’s chemical industry because they can obtain precise experiment results automatically and hence allow lab technicians to focus on other tasks concurrently. One time consuming process is the process of sampling chemical test results into the correct container, especially when the tests are many and infrequent. This is where an automated sampling machine, or an autosampler, comes into play. If this sampling process is automated, the accurate sampling is achievable without the full time attendance of a lab technician. This thesis explores how to construct a small footprint autosampler in the form of a gantry, complete with electronics, actuation of all axes and relevant embedded software needed for implementation in a lab setting. This thesis also explores how to optimally control the stepper motors in charge of axis actuation, by comparing classic open loop control, PD control and an adaptive LQI controller. This is done by developing a mathematical model of the system using MATLAB Simulink, followed by developing and optimizing the controllers on the model and then implementing and verifying the controllers on the physical system. The conclusion of this thesis is that the PD controller performs best for controlling a stepper motor. It performs slightly better than open loop control and has the additional advantage of reference following. The adaptive LQI performed the worst on the model and additionally, it could not be implemented on the physical system due to how the system was constructed in hardware.

NoviOcean is a company dedicated to the development of wave power plants. In order to efectively demonstrate and explain the functionality of their construction to investors and clients, they sought to create a small prototype of the wave power plant capable of movement. The purpose of this report was to examine the feasibility of replicating ocean waves using a mechanical approach by constructing a Stewart Platform for NoviOcean. A Stewart Platform consists of 6 linked legs connected by a top and a base plate. By adjusting the length of the legs, the top plate can move in 6 degrees of freedom. Depending on the angle of the servo motors, the length of each leg changes, thus afecting the position of the top plate. The length of each leg is determined using inverse kinematics. The fnal product can move like a particle in the ocean, in a circular path, with satisfactory tolerances. However, there are visible tremors in the movement, likely due to material and manufacturing defciencies. Further work is needed to address this issue.

Human activity data is typically represented as time series data, and RNNs, often with LSTM cells, are commonly used for recognition in this field. However, RNNs and LSTM-RNNs are often too resource-intensive for real-time applications on resource constrained devices, making them unsuitable. This thesis project is carried out at Wrlds AB, Stockholm. At Wrlds, all machine learning is run in the cloud, but they have been attempting to run their AI algorithms on their embedded devices. The main task of this project was to investigate alternative network structures to minimize the size of the networks to be used on human activity data. This thesis investigates the use of Fast GRNN, a deep learning algorithm developed by Microsoft researchers, to classify human activity on resource-constrained devices. The FastGRNN algorithm was compared to state-of-the-art RNNs, LSTM, GRU, and Simple RNN in terms of accuracy, classification time, memory usage, and energy consumption. This research is limited to implementing the FastRNN algorithm on Nordic SoCs using their SDK and TensorFlow Lite Micro. The result of this thesis shows that the proposed network has similar performance as LSTM networks in terms of accuracy while being both considerably smaller and faster, making it a promising solution for human activity recognition on embedded devices with limited computational resources and merits further investigation.

Four-dimensional (4D) printing is an advanced manufacturing technology that has rapidly emerged as a transformative tool with the capacity to reshape various research domains and industries. Distinguished by its integration of time as a dimension, 4D printing allows objects to dynamically respond to external stimuli, setting it apart from conventional 3D printing. This roadmap has been devised, by contributions of 44 active researchers in this field from 32 affiliations world-wide, to navigate the swiftly evolving landscape of 4D printing, consolidating recent advancements and making them accessible to experts across diverse fields, ranging from biomedicine to aerospace, textiles to electronics. The roadmap's goal is to empower both experts and enthusiasts, facilitating the exploitation of 4D printing's transformative potential to create intelligent, adaptive objects that are not only feasible but readily attainable. By addressing current and future challenges and proposing advancements in science and technology, it sets the stage for revolutionary progress in numerous industries, positioning 4D printing as a transformative tool for the future.

This paper presents the development and evaluation of a prototype for medication management, developed for individuals with Attention Deficit Hyperactivity Disorder (ADHD). The research methodology implemented included sensor testing, prototype construction, user trial, and analysis of the results. Ultrasonic and IR sensors were tested and evaluated regarding their detection range, closeness, and performance with different materials and lighting. Due to a similar results between the two sensors and a low validity, the IR sensor was picked due to its size being more compatible with the prototype. Further the prototypes were developed and tested by eight people with ADHD, with the iteration of the second prototype being based on their feedback. Feedback from the interviews showed a positive response to the prototype’s features, including its customizable LED lights and ability to track the amount of medication taken. The user group rated eighth aspects and the average rating increased for six of them on the second prototype. The findings from this study show a positive ability for technical solutions made for individuals with ADHD, emphasizing the importance of user-centered design and customization options in health technology interventions.

The state-of-the-art logic elements, like CPUs and FPGAs, are either trending towards higher core counts or heterogeneous architecture, driven primarily by the telecommunication and the automotive industry. The railway industry market, although relatively small, adopts logic elements from these leading sectors. However, there are significant challenges in adopting logic elements from other domains for the railway signaling systems which are expected to operate continuously for nearly thirty years ensuring the highest levels of safety. Furthermore, the intricacies involved in demonstrating safety prevent signaling applications from fully harnessing the capabilities of multi-core or heterogeneous architectures. Our paper focuses on the utilization constraints and challenges of using the latest logic elements in signaling applications conforming to EN 50129:2018. We analyze the use of such logic elements and architectural patterns to meet safety requirements cost-efficiently.

This study investigated the implementation of a remaining range estimation algorithm for electrical vehicles, an essential feature to define a vehicle's reliability on the road. The implementation was made on an electrical motorcycle, comparing three models: a dynamic force based model, a power based model and a mass estimation model. The mass model estimated the mass with the help of a RLS algorithm and is a combination of the force based model and the power model. It investigates the possibility to further increase the accuracy of a range estimation algorithm by estimating the total mass of the vehicle over a driving session. On top of these models, two kinds of prediction methods for future consumption were evaluated: the average-past prediction and the home-intention prediction. Both models uses past data to predict the future, but the home-intention prediction is a suggested method to further improve the classic average-past method, where the beginning and end of the vehicle's driving sessions is assumed to be the same location. Tests were executed for the models on an electrical motorcycle provided by the company CAKE. A test equipment were put on the motorcycle, consisting of microprocessors and sensors, used for computation and collection of data. With this equipment, experiments were performed on three test routes with different conditions, comparing the models’ accuracies. The results showed that the Power Model, even with its lower complexity performed best overall, while the Force Model showed mixed results. Depending on the prediction method the Force Model performed either at the top or at the bottom. When the results were analyzed, this behavior seem to be the result of insufficient/faulty hardware which were essential for the average-past prediction to achieve proper results. The Force Model using home-intention prediction consistently performed better, as long as its prediction was correct. The Mass Model was executed offline and were then used to simulate the effect it could have had online. This showed promising result, suggesting improved accuracy if implemented online, but which in this thesis is left as a suggestion of improvement for future work.

This Bachelor’s Thesis examines the construction, implementation, and control of a conventional disc brake using an electric motor to achieve its braking force. As self-driving, automation, and AI increasingly become the norm in the automobile industry, it becomes more and more interesting to look at diferent areas where improvements can be made. While hydraulics remain the preferable choice in the automotive industry, in some applications an electric system could be preferable. For this Bachelor’s Thesis, we focused on the disc brake, creating a functional prototype that uses an electric motor to achieve its braking force and then controlling that motor to facilitate speed regulation, anti-lock braking, and incremental braking force. The prototype was made using mostly 3D-printed parts with some laser-cut acrylic and utilized a mechatronic system to measure and control the speed as well as the braking force. The prototype was largely successful, achieving all its goals. However, these goals were not implemented to a satisfactory degree in its current state. With further development, the prototype could be useful in micro-mobility applications or other smaller electronic appliances. Furthermore, this project provided important insight into how modern friction brakes can be customized with electronics.

The concept of Digital Twins has attracted more and more attention from both the academia and the industry in the recent years, much thanks to the possibility of bidirectional communication between physical objects and their digital replicas. However, despite its growing prominence, the existing literature on Digital Twins remains limited. This is especially true for literature on guidelines and frameworks for constructing digital models for use in Digital Twins. This thesis examines the kinematics underlying such digital models, aiming to create an optimized kinematic solution for the pick-and-place robot Codian Delta D4. More specifically, the report presents a comparative case study to investigate how two different digital models, characterized by varying levels of kinematic detail, impact the performance of the Delta D4 robot in terms of accuracy and computational efficiency. The results reveal that the inclusion of additional details did not significantly impact computational efficiency. However, it was observed that the added details introduced minor deviations in accuracy. A continued assessment is therefore necessary to decide which deviations are acceptable for the model’s intended purpose and whether the resources required for the extra development are justified. The authors suggests that the more detailed model, excluding the spring components, is a promising foundation for a future implementation of a Digital Twin of the Delta D4 robot.

In this thesis, a catapult inspired robot is constructed which can aim and throw small wooden balls at targets. The goal was to achieve the best possible accuracy, precision and throwing capability. The robot was mainly constructed using 3D printers, where a stepper motor was used for the aiming mechanism and a high torque DC-motor was used for the throwing mechanism. All electrical components were connected to an Arduino Uno microcontroller, from which all the needed calculations were made. When tested, the robot could throw a maximum of 4.8 m with a wooden ball with mass 7 g. The precision was high but the accuracy was below expectations and varied depending on the distance and direction of the target. The robot would hit its targets 66% of the times, where the targets consisted of horizontal discs with radius 7.5 cm.

Automated driving vehicles have shown glorious potential in the near future market due to the high safety and convenience for drivers and passengers. Image sensors' reliability attract many researchers' interests as many image sensors are used in autonomous vehicles. We propose an online image sensor fault detection method based on comparing the historical variances of normal pixels and defective pixels to detect faults. For fault pixels without uncertainty, with a detecting window of more than 30 frames, we get 100% accuracy and 100% recall on realistic continuous traffic pictures from the KITTI data set. We also explore the influence of fault pixel values' uncertainty from 0% to 25% and study different fixed thresholds and a dynamic threshold for judgments. Strict threshold, which is 0.1, has a high accuracy (99.16%) but has a low recall (34.46%) for 15% uncertainty. Loose threshold, which is 0.3, has a relatively high recall (83.78%) but mistakes too many normal pixels with 18.17% accuracy for 15% uncertainty. Our dynamic threshold balances the accuracy and recall. It gets 100% accuracy and 58.69% recall for 5% uncertainty and 78.38% accuracy and 55.39% recall for 15% uncertainty. Based on the detected damage pixel rate, we develop a health score for evaluating the image sensor system intuitively. It can also be helpful for making decision about replacing cameras.

The purpose of this project was to create an autonomous plant watering system for potted plants to answer the research questions. Having healthy plants at home has proven to be beneficial by improving air quality and creating a peaceful environment, therefore reducing stress. The focus of this research paper was to study how autonomous the system can be, and how the important parameters are regulated. To answer our research questions a system was created and tested. The system used a microcontroller and a handful of sensors together with actuators, including pumps and an LED strip. The goal was to regulate the pH and nutrients of the water supplied to the plant, moisture level in the plants soil and the light that the plant received. Tests were conducted on a potted basil in the mechatronics lab at KTH. Results from experiments showed that our system for lighting worked as intended, however the conversion from analog value to lux was faulty due to the usage of default values for calibration. pH regulation works, but takes over 4 hours to stabilize, making the system less optimal. Lastly, due to the extensive time it takes for the water container to be emptied by watering the plant, calculations were made from the data gathered during the 228-hour long moisture monitoring experiment. This, together with research and further calculations, showed that the container needs to be refilled roughly every 11 weeks, making it autonomous to a high degree.

This thesis investigates the feasibility of using a PID-controlled reaction wheel to stabilise an otherwise unstable miniature bike. Although such a prototype is not easily comparable to a full-sized bike, the implications and challenges that this would propose could still prove to be relevant. Through a series of tests, the performance of the controller is examined, assessing its capabilities both in isolated conditions, and scenarios where disturbances are introduced. The results reveal a nuanced understanding: while the PID-implementation offers a foundational framework, its standalone application often falls short on this scale. Because of this, supporting functionalities that shifts the reference point of the controller are explored to improve the system’s capabilities. This proves to be an effective method, enabling the bike to balance much more rigorously. However, despite the bicycle’s ability to maintain balance indefinitely, there still exist areas of improvement, including enhancing the system’s responsiveness, dampening sensor noise, and refining the design of the prototype. By optimising factors like these, an attempt at constructing a fullsized bicycle may be possible, proposing a safer way of travel.

As mechatronics is a more and more relevant part of the modern society, new inventions that can ease everyday life are constantly developed. The idea of this project was to design and build a prototype of a self balancing bicycle. A few examples of areas of applications for the bicycle could be package delivering services, transport vehicles for people that have reduced cognitive functions, warehouse work or nature surveillance. The prototype robot was designed as a small-scaled bicycle with a reaction wheel attached to an electrical motor. The balancing act was performed by measuring the offset angle from equilibrium with a sensor equipped with an accelerometer and a gyroscope, which communicated with the motor. All of the components were connected to an Arduino Uno, which allowed for the collection of data from the sensor readings, as well as control of the motor. A PID controller was implemented in order for the prototype to achieve the desired balancing behaviour. The functionality of the components were tested extensively, both individually and integrated together in the system. The results were that the robot performed the correct motion based on the sensor readings but due to the PID controller not being tested enough, the bicycle did not manage to balance on its own. The conclusion was that this type of robot can indeed be built and function as intended but that more fine tuning of the controller would be required for the bicycle to balance on its own.

The purpose of this project is to examine how a car controlled by hand gestures wirelessly can be constructed. The hand gestures are measured with a gyro sensor. This project examines how the time delay can be reduced with focus on data transmissions protocols and code structure. Furthermore, the steering is examined so the car can turn while driving. The choice of transmitter was made regarding the latency where nRF24L01 had the lowest latency. To reduce the time delay, the blocking function delay() was avoided which pauses the entire program and instead the millis() function was used, which instead is used to specify how often a code block should run without blocking the rest of the program. Therefore, the Arduino program does not pause, and new gyro signals can be received continually. The time delay was 2.64 ms. The steering is controlled with a servo which is connected to a four-bar linkage inspired by Ackerman’s steering principle. To examine the steering the angles of the servo and the front wheels were measured at different angles of the hand.

Robotic arms and similar technology are widely present within manufacturing today, increasing productivity massively. As the field of robotics progresses and artificial intelligence becomes more refined, it is likely we will see further applications of these technologies that may extend into daily life. This bachelor’s thesis will cover the development of a robotic arm, capable of interacting with a person through playing a game of Tic Tac Toe. The robot will be constructed with three degrees of freedom, wherein each joint will be driven by a servo motor. The playing field will be monitored by a camera mounted onto a simple tripod. It will identify the game pieces by analysing their RGB-values and mapping them onto a digital board accordingly. A minimax algorithm will then determine the best move to make and convert that into a set of coordinates on the board. Finally, those coordinates can be calculated into joint angles through inverse kinematics. Results show that the movement is precise with an accuracy of above 80%, games that are played with consistent lighting have a completion rate of 70% and average move time is 21.55 seconds.

This bachelor’s thesis explores the stabilization of an inherently unstable system, the inverted pendulum, using a Proportional-Integral-Derivative (PID) controller implemented in a simulated environment and tested in hardware. The inverted pendulum poses a classic control problem due to its natural tendency to fall over, making it an ideal project for exploring different control systems. To solve the problem of stabilization of the pendulum a physical model of the system was constructed and the dynamics of the system were modeled in the Simulink environment. A PID regulator was chosen as the controller and different methods were used to find PID coefficients for achieving system stability. Various experiments were conducted to fine tune the gain values. The results demonstrate that while the system could be stabilized for a short period of time, maintaining long term stability proved challenging due to mechanical limitations and the non-ideal behaviors of the components, particularly the stepper motor used for actuation. These experiments provided insights into the discrepancies between simulated predictions and actual performance, highlighting the impact of factors choosing right input values for the simulation and mechanical limitations. Ultimately, this project reinforced the theoretical foundations of control systems and highlighted practical challenges that are present in such a project, also offering improvements for alternative actuation mechanisms and more robust and better suited control algorithms.

Today, prosthetics are an essential tool for amputees who want to live a normal everyday life. In order to further our knowledge within the area of prosthetics, we decided to create a robotic hand that closely resembles the human hand in its anatomy and function. In order to control and test the prosthesis we decided to create a medium that can recognize human motion and recreate it in the mechanical hand as accurately as possible. Our challenge would be to design an artificial copy of an organic mechanism using known mechanical principles, as well as to find a suitable way to capture physical movements made by a person and make them into instructions for the mechanical hand. To achieve this, we decided to attach flex sensors to a glove so that a microcontroller could read their tension to control the arm. We used Solid Edge 2023 to design the individual parts of the arm, based on real human anatomy and we used fishing wire attached to servo motors to control the fingers. The glove and hand were connected by code that translated the flex sensor signals into control signals for the servo motors. In the end, we managed to create all parts of the project and the robotic hand was able to hold small objects for longer than 30 seconds. Although it would have been preferable if the hand could hold heavier objects the project is seen as a success, since we managed to further our knowledge in anatomy, electronics and prosthetics.

One of the major challenges facing society today is efective waste management, making recycling initiatives crucial for reducing environmental pollution and conserving valuable resources. This thesis aims to develop an autonomous system capable of searching for and collecting discarded recyclable red cans in parks and other public areas. The paper presents the hardware design, software development and testing results of a prototype created for this purpose. The hardware setup includes movement through DC motors and a pickup process involving servo motors, which facilitates the collection of cans into a container mounted on the prototype. The software component uses object detection, utilizing image processing techniques to recognize and locate red cans within the environment. A control system integrates image processing, sensor feedback and wheel control to navigate and maintain the correct position in front of the target cans. Testing results reveal a success rate of 80% and a well-functioning control system. However, challenges in object detection under extreme lighting conditions, as well as limitations in the control system’s precision, particularly in sensor reliability, were observed. The discussion highlights the need for further development, including the implementation of machine learning models for object recognition and enhancements to the retrieval process for improved effciency.

Stateflow models are complex software models, often used as part of industrial safety-critical software solutions designed with Matlab Simulink. Being part of safety-critical solutions, these models require the application of rigorous verification techniques for assuring their correctness. In this paper, we propose a refutation-based formal verification approach for analyzing Stateflow models against invariant properties, based on bounded model checking (BMC). The crux of our technique is: i) a representation of the state space of Stateflow models as a symbolic transition system (STS) over the symbolic configurations of the model, and ii) application of incremental BMC, to generate verification results after each unrolling of the next-state relation of the transition system. To this end, we develop a symbolic structural operational semantics (SSOS) for Stateflow, starting from an existing structural operational semantics (SOS), and show the preservation of invariant properties between the two. We define bounded invariant checking for STS over symbolic configurations as a satisfiability problem. We develop an automated procedure for generating the initial and next-state predicates of the STS, and a prototype implementation of the technique in the form of a tool utilising standard, off-the-shelf satisfiability solvers. Finally, we present preliminary performance results by applying our tool on an illustrative example and two industrial models. 

The complexity of deploying automated vehicles (AVs) has been grossly underestimated and vehicles at high levels of automated driving (SAE level 4 and above) have so far only been deployed in very limited areas. Highly automated AVs will face complex traffic, e.g., due to occlusions and unpredictable road-user behaviour, and AVs may shift the distribution of crashes. This has given rise to a renewed interest in connectivity and collaboration, with the need to monitor (emerging) behaviours and risk, and the promise to improve road traffic safety and performance by resolving the “information gap”. This motivates further investigations and research in this direction. In this paper we set out to identify barriers and important directions towards solutions for such collaborative systems, as formed by connected automated vehicles and a supporting cyber-physical infrastructure. Drawing upon a state-of-the art assessment and interactions with experts, we conclude that the current state-of-the art is fragmented, and therefore investigate key topics related to collaboration barriers and propose research questions to address them, hoping that the provided structure can also assist in guiding future research. The topics cover, (i) the socio-technical and system of systems nature of collaborative systems, (ii) the multifaceted design space and architectures with related trade-for such systems including between safety, performance and cost, and (iii) trustworthiness issues, ranging from safety and cybersecurity to privacy and ethics.

This project explores the development of a prototype for a line-following robot equipped with mecanum wheels, aiming to combine omnidirectional movement with precise path-following capabilities. The robot is powered by stepper motors and uses sensors for color recognition to identify paths and directions. A microcontroller manages the control system, implemented using a PID algorithm for accurate speed regulation. The research analyzes the feasibility of integrating mecanum wheels and automated path detection through sensors. Challenges encountered include slippage during turns and sensor calibration issues, providing valuable insights into the limitations and advantages of this technology for autonomous guided vehicles. Future improvements should focus on scaling up the motors and enhancing the sensor system for better line detection and refining the control algorithms for smoother directional changes.

Research background: Digital ecosystems in Europe are heterogenous organizations involving different economies, industries, and contexts. Among them, Digital Innovation Hubs (DIHs) are considered a policy-driven organization fostered by the European Commission to push companies’ digital transition through a wide portfolio of supporting services.

Purpose of the article: There are DIHs existing in all European economies, but literature needs more precise indications about their status and nature. The purpose is to study a distribution of DIHs and differences in portfolios of DIHs’ services across European economies. Therefore, the paper wants to deliver more precise data on effects on national and European policies. This is required to define their final role and scope in the complex dynamics of the digital transition, depending on regional context and heterogeneity of industries.

Methods: Data on 38 economies was collected from the S3 platform (on both existing and in preparation DIHs) and further verified by native speaking researchers using manual web scrapping of websites of DIHs identified from S3. To find potential similarities of digital ecosystems in different economies as emanated by the existence of DIHs, clusterization (Ward’s method and Euclidean distances) was applied according to the services offered. Economies were clustered according to the number of DIHs and the spread of DIHs intensity in different cities. The results were further analyzed according to the scope of the provided services.

Findings & value added: The applied clustering classified European economies in four different sets, according to the types of services offered by the DIHs. These sets are expression of the different digitalization statuses and strategies of the selected economies and, as such, the services a company can benefit from in a specific economy. Potential development-related reasons behind the data-driven clustering are then conjectured and reported, to guide companies and policy makers in their digitalization strategies.

This thesis documents the construction and programming of an autonomous robot capable of playing checkers. The robot will physically play against a human player, by interacting on a common playing field. It will register the player’s actions on the playing field using an RGB-sensor and react according to the rules of checkers. Although the robot will be programmed in arduino c++, the project will implement an existing minimax algorithm from python to perform the calculations and decide on the robot’s next move. The python script will be implemented with the help of serial communication inorder to control the arduino. The project resulted in a final prototype capable of playing a full game of checkers. However, improvements can still be made in terms of its efficiency, primarily to the algorithm.

In this chapter, we propose a roadmap for the development of education in mechatronics. This development is intended to support future mechatronics engineers in accepting challenges in evolution, development, and sustainability. We explore curricula design, learning strategies, and misalignment between education and professional needs in relation to these challenges. The exploration is based on the fact that the subject mechatronics has now been around for 50 years, and, in particular, relates to the experiences and previous research of the authors, who are based at KTH Royal Institute of Technology where most of the empirical data is from. In this chapter, we argue that the above challenges require new roles, skills, and competencies among mechatronics engineers. With examples from complex product development, we argue that an increased systems perspective with systems architect competencies and experiences can support the embracing of additional systems design factors, which, for example, can enable sustainable development aspects into product design. We explore the interface between mechatronics and systems engineering, via embedded systems, as a possible future development of the subject of mechatronics. Embedded systems can be seen as a related subject, or as a subset of mechatronics, or vice versa. With this chapter, we argue that the subject of embedded systems can provide a shortcut for mechatronics education to approach the role of systems engineering, thereby reaching a level of competence and skills where the above challenges can be approached. The mechatronics engineer can play an important role in sustainable development. With existing curricula, mechatronics engineers are prepared to accept complex problem-solving, utilize solutions from various domains and disciplines, and create domain-independent solutions. These skills are crucial in embracing sustainable development, which puts the mechatronics engineer in a unique position. We also bridge the above discussion with current trends in learning, such as life-long learning and e-learning. The mechanisms that enable the mechatronics engineer to achieve a systems engineering skillset, which, according to our results, needs to be based on engineering experiences, is facilitated both by a professional education system (life-long learning) and by a flexible setup (e-learning).

TECoSA – a university-based research center in collaboration with industry – was established early in 2020, focusing on Trustworthy Edge Computing Systems and Applications. This article summarizes and assesses the current trends and drivers regarding edge computing. In our analysis, edge computing provided by mobile network operators will be the initial dominating form of this new computing paradigm for the coming decade. These insights form the basis for the research agenda of the TECoSA center, highlighting more advanced use cases, including AR/VR/Cognitive Assistance, cyber-physical systems, and distributed machine learning. The article further elaborates on the identified strategic directions given these trends, emphasizing testbeds and collaborative multidisciplinary research.

Combining machine learning and formal methods (FMs) provides a possible solution to overcome the safety issue of autonomous driving (AD) vehicles. However, there are gaps to be bridged before this combination becomes practically applicable and useful. In an attempt to facilitate researchers in both FMs and AD areas, this paper proposes a framework that combines two well-known tools, namely CommonRoad and UPPAAL. On the one hand, CommonRoad can be enhanced by the rigorous semantics of models in UPPAAL, which enables a systematic and comprehensive understanding of the AD system’s behaviour and thus strengthens the safety of the system. On the other hand, controllers synthesised by UPPAAL can be visualised by CommonRoad in real-world road networks, which facilitates AD vehicle designers greatly adopting formal models in system design. In this framework, we provide automatic model conversions between CommonRoad and UPPAAL. Therefore, users only need to program in Python and the framework takes care of the formal models, learning, and verification in the backend. We perform experiments to demonstrate the applicability of our framework in various AD scenarios, discuss the advantages of solving motion planning in our framework, and show the scalability limit and possible solutions.

Robotic arms have a wide range of uses, for example in the industry. They can be used to automate and standardize manufacturing processes et cetera. Many of these robotic arms are very complex and have many degrees of freedom. The robotic arm that has been manufactured in this project is very different both in design and application. The arm itself has only one degree of freedom similar to a catapult. The purpose of the robot is to play Beer Pong, that is, throw ping pong balls into plastic cups. What is primarily explored in the project is the precision of the robot's throw. The key components are a camera to detect plastic cups with color recognition, as well as measure how far away the cup is. Two stepper motors are used, one to drive the throw and one to rotate the structure. A claw is also used to grip and release the ball, which is powered by a servo motor. All motors are controlled via an Arduino UNO, and control of the camera and all calculations are done in Python on a laptop. When testing the robotic arm, 37% of all throws hit, although about 80% of all throws were very close to hitting and hit the edge of the cup. The robot's maximum throw length is around 60 cm.

Robotic lawnmowers have become increasingly ubiquitous in green spaces worldwide, offering automated lawn maintenance solutions. However, a prevalent issue with many existing models is their reliance on randomized pathing algorithms rather than optimized navigation. This report addresses this challenge by presenting the design and implementation of an Arduinobased robotic lawn mower featuring dual control systems: an autonomous mode driven by a sensor suite and a manual mode controlled via a handheld device. The report will answer the question if the additional manual control system proves to be an effective way to control a robotic lawn mower as well as how the prototype compares to current lawn mowers on the consumer market. To answer these questions a prototype was designed and assembled using cheap and easily available parts such as an Arduino microcontroller and 3D-printed parts. The sensor suite was designed utilizing a combination of custom-built sensors and commercially available off-the-shelf sensors. The controller was constructed using an Arduino microcontroller in conjunction with an NRF communication module. Based on conducted results, the manual override control demonstrated a more efficient operational time compared to autonomous randomized path planning. All though the manual control system proved to be an effective way to control the mower, it can be time-consuming, particularly in larger or more complex lawn environments. It proved to be ineffective in the sense that the constant supervision needed defeated any time gains made.

The implementation of automated systems in manufacturing industries increases efficiency, precision, and safety by reducing human intervention, errors, and waste. Variable loads can cause several problems for automation systems. One of the most significant challenges is maintaining the stability and precision of the production process despite changing load conditions. These variable loads can lead to unstable systems or failures, causing an increase in errors, reduced efficiency, and lower product quality. It is essential to design control systems that can adapt to changing load conditions and maintain stable and precise operation under all circumstances. To address this problem, this thesis presents an adaptive controller based on load identification and gain scheduling, to replace the standard FOC consisting of regular PI-controllers. The load estimator is used to estimate the external load with relatively small RMSD values, while the ain scheduler adjusts the controller gains based on the estimated load. Other controllers are also explored, such as an angular velocity error-based adaptive controller. The results shows that both proposed controllers perform better than the standard controller when the system is subject to variable external loads, however, the load estimator paired with the gain scheduled PI-controller performs best.  

We present the integration of a Non-Axiomatic Reasoning System (NARS) with mobile robots for planning and decision making. NARS enables robots to effectively handle uncertainty in real-time with complete sensor and actuator integration, thereby ensuring adaptability to evolving scenarios. We discuss essential parts of the logic, the architecture and working principles of NARS, and the integration of NARS as a ROS node. A case study is provided demonstrating the system's proficiency to carry out a garbage collection task in an open-air environment by operating a mobile robot with manipulator arm, and we demonstrate its ability to learn about the place-dependent accumulation of garbage items. Case study also reveals that our approach performs more effectively on the overall task than the Belief-Desire-Intention model we compared with.

In industrial-sized cyber-physical systems, ensuring fulfillment of requirements gets increasingly more costly as the number of components increases. To make the task feasible, compositional verification has been suggested as a scalable solution. Such techniques allow verification by divide-and-conquer, often using assume-guarantee contracts. Although previous research has focused mostly on the non-probabilistic setting, in the real world, probabilities often arise due to random hardware failures, stochastic communication delays, sensor ghost objects, machine learning components, rounding errors caused by finite-precision arithmetic, human behavior, and probabilistic algorithms. Therefore, for contract theories to be practically relevant to cyber-physical systems, there is a need to support probabilistic reasoning, for instance regarding safety and reliability. To this end, we propose a completely trace-based probabilistic contract theory, supporting general probability measures, continuous time, and continuous state spaces. To verify decompositions of such contracts, we also present a deductive system, which is illustrated on an industrially inspired automatic emergency braking example.

Compositional verification allows a system to be verified indirectly by verifying the individual components of the system. The keystep is to ensure that the decomposition of the system specification intocomponent specifications is correct. That is, it needs to be verified thatthe composition of the component specifications refines the system specification. In many cyber-physical systems, specifications are probabilisticin nature. For instance, a specification might state that the probabilityof reaching an unsafe state within 10.000 hours shall be less than 0.05.Verifying refinement under such assumptions requires techniques beyondtraditional theorem proving. This paper presents a solution where specifications are built up by probabilistic contracts based upon timed automata. In particular, the main contribution is an algorithm for verifyingrefinement between such specifications. The algorithm utilizes a reduction to the language emptiness problem, making the algorithm terminateafter a finite number of computations

This project aimed to develop an automatic drink cooler and dispenser prototype utilizing a coolant mixture based on ice and salt. The main targets for real-life applications are settings such as bars and festivals where electricity is limited and the demand for creating cooled drinks without staff is high. The challenge consisted of getting the temperature in the cooling bath to convert to a time for which the beverage should be cooled for. The construction consisted of a Arduino UNO microcontroller, two Brushless DC-pumps, one servo motor powering a valve, two liquid containers, and a cooling chamber. Experimental results indicated that the prototype achieved satisfactory performance. It proved possible to calculate the duration for which the beverage should remain in the cooling chamber by using the temperature of the cooling bath. From the data, two predictive equations were formulated based on different desired temperatures of the drink. For a target beverage temperature of 8◦C, the model was t = 8.9x + 132.9, and for a target beverage temperature of 6◦C, the model was t = 10.3x + 191.7. For both models t represents the cooling time and x represents the temperature of the cooling bath. These models facilitate precise control within an acceptable error margin of ±15%, enhancing the prototype’s possibility to be utilized in the intended environments.

Nucleic acid amplification tests (NAATs) are powerful medical diagnostic tools for point-of-care (POC) and other field applications. However, traditional methods like qquantitative PCR (qPCR) require complex, expensive equipment and trained operators, limiting their use to centralized labs. Isothermal alternatives, like Loop-mediated Isothermal Amplification (LAMP), are better adapted for POC devices. Lab-on-PCB systems have the potential to overcome the challenges faced by conventional microfabrication-based systems. This study presents a novel lab-on-PCB device for RNA amplification and electrochemical detection using reverse transcription LAMP (RT-LAMP) of SARS-CoV-2. The system consists of only two disposable PCB-based chips making it close to zero cost. One PCB is for heating and DNA amplification, while the other is for electrochemical detection using Cyclic Voltammetry with a redox-active intercalating probe. The PCB slides are connected to a compact electronic device (<10 USD) for controlling the heating and electroanalytical readout. Using this device, we achieved successful rapid (<1 hour) nucleic amplification and detection at a target concentration of 100 copies/reaction. This work represents a notable step toward developing integrated, portable NAAT devices for POC diagnostics.

Our world contains advanced constructions such as towering skyscrapers, intricate infrastructure, cargo ships and spacecraft. The inspection and maintenance of these constructions at great heights can expose human workers to severe risks. A solution to the risks posed to humans could be the usage of wall climbing robots.The aim of this bachelor thesis is therefore to construct and test a wall climbing robot using electromagnetism as its adhesive method. The goal was to create a robot that could climb in a straight line on a ferromagnetic planar surface and also refrain from going outside the edges of the wall. In addition, the robot should also be able to climb on a wet wall. The robot frame was constructed with a main body and four moving limbs. The robots ability was evaluated with three tests corresponding with the projects goals. The results showed the robot had a difficult time climbing in a straight pattern. This was caused by it’s geometry and lack of mobility in the limbs. The implementation of rails helped with the ability to move forward, but made the movement pattern unpredictable. This was caused by the electromagnets sliding up and down within the railing. Detecting the edges of the wall was not a problem and could be solved with different distance measuring sensors. Applying water to the wall’s surface made the robot slower and prone to slippage. However, it did not make it unable to move or made it slide off the wall.

In the modern day of environmental consciousness and energy optimisation, exploring the possibilities of implementing obstacle avoidance designs may be relevant to possibly streamline energy use in cars and road vehicles. This B.Sc thesis project covers the design and construction of a prototype vehicle made for testing simple obstacle detection and avoidance cases. By mounting the vehicle’s wheels using extension units it is possible to raise and lower them at will and thus evade hurdles. Previous work on the subject has concluded the solution to be possible, yet inadequate due to the toppling effect when balancing a car on three wheels. In this project a mass relocating system is designed and implemented to counteract this effect. It was found that the prototype is able to detect and traverse an obstacle with a height of 90 millimeters, which is almost three times greater than the radius of the wheels. However, the clearance was often made at an insufficient distance to the obstacle. The mass relocating system is adequate for keeping the prototype in balance and the assembled prototype is successful in traversing obstacles. However it is in many cases unable to detect the obstacle at a sufficient distance to avoid contact with the front wheels.

This bachelor’s thesis is based on the empirical construction of a line follower robot, upon which testing of PID feedback control systems was done. Line following as a subject is mainly applied in warehouses and autonomous driving of road vehicles, as well as in some trivial competitions. As the name would suggest, a line follower robot aims to follow a path determined by a line as accurately as possible, which is where feedback control systems come into play. Additionally in our case, the aim was to (with the use of encoders) keep the average speed of the robot consistent despite drops in voltage due to various factors such as overheating. The robot was designed and constructed using various machines and material available, and then programmed using an Arduino Uno. The resulting line follower was significantly improved by the PID controller tuned by the Ziegler Nichols method. Additionally, applying a second PID controller for the encoders to ensure constant average speed not only worked, but also improved the accuracy of the line follower. Lastly, the attempt at systematically improving the PID constants of the line follower successfully yielded a 8.5% reduction in accumulated error.

Transitioning the transport sector to clean energy sources is crucial for mitigating greenhouse gas emissions and achieving carbon neutrality. A collaborative solution, combining both electric vehicles and combustion engines using renewable fuels, may prove more effective than competitive ones. This necessitates a focus on developing sustainable combustion engines by improving their efficiency through renewable energy sources and innovative technologies.

This thesis uses exergy analysis to evaluate engine efficiency, losses, and irreversibilities, as well as the work potential of exhaust flows. Particular emphasis is placed on the implications of these exergy analyses in relation to engine operations, especially concerning combustion processes and exhaust pulsations. Exergy analysis quantifies the maximum work extractable from an energy source, enabling the identification and quantification of losses and inefficiencies in thermal processes. A dual-fuel marine engine with two-stage turbocharging and an ethanol-fueled heavy-duty spark-ignition (SI) engine using lean burn are examined with validated one-dimensional engine models to analyze engine performance and losses from an exergy perspective. In the tested marine engine, irreversibilities are quantified and categorized into three types, with combustion irreversibility being the most significant, followed by losses through gas exchange and heat dissipation. In the ethanol-fueled SI engine, the effect of lean-burn combustion at high load is investigated through the excess air ratio up to 1.8, assessing its impact on thermal efficiency, combustion phasing, as well as energy and exergy distributions. Results indicate that employing lean burn improves engine efficiency with advanced combustion phasing but also leads to more exergy destruction. The importance of maintaining high exergy recovery through turbocharging for diluted operation is also highlighted.

Additionally, high-frequency exhaust pulsations resulting from valve motion pose challenges in accurately resolving exhaust energy and exergy. To address this, this thesis investigates methods for exhaust pulse characterization and measurement under unsteady flow conditions. Sensitivity analyses, based on a heavy-duty engine simulation, highlight the importance of time-resolved mass flow measurements in quantifying the energy and exergy of exhaust pulsations. Subsequently, this research implements a Pitot tube-based approach to measure crank angle-resolved engine exhaust mass flow rates and to further analyze the effect of attenuated temperature measurements on resolving instantaneous mass flows. The findings indicate that temperature variations pertaining to exhaust flow conditions have only a relatively small impact on mass flow measurements. Based on the exhaust flow measurements, the mass flow characteristics of exhaust pulsations are also discussed with regard to the blow-down and scavenge phases.