The ambition of this project was to create, program and control a matrix of individually controllable LEDs. This was achieved by utilizing a variety of microcomputers, such as Arduino, and by forging the hardware required for signal processing. Several subsystems were constructed, a main unit for controlling the LEDs, a signal processing unit for analyzing audio signals by employing the Fourier transform, and a stage lighting communication unit to communicate with standard stage equipment. All the subsystems were conjoined into one complete system using digital communication. The physical construction of the matrix was not finished but the subsystems work together and are able to control a miniature prototype of the matrix (with the same characteristics as the final build) with high responsiveness and precision.

In the last decade, socially assistive robots have been used in therapeutic treatments for individuals diagnosed with Autism Spectrum Disorders (ASDs). Preliminary studies have demonstrated positive results using the Penguin for Autism Behavioral Intervention (PABI) developed by the AIM Lab at WPI to assist individuals diagnosed with ASDs in Applied Behavioral Analysis (ABA) therapy treatments. In recent years, power-efficient embedded AI computing devices have emerged as a powerful technology by reducing the complexity of the hardware platforms while providing support for parallel models of computation. This new hardware architecture seems to be an important step in the improvement of socially assistive robots in ABA therapy. In this thesis, we explore the use of a power-efficient embedded AI computing device and pre-trained deep learning models to improve PABI’s performance. Five main contributions are made in this work. First, a robot-enhanced ABA therapy framework is designed. Second, a multilayer pattern software architecture for a robot-enhanced ABA therapy framework is explored. Third, a multifactorial experiment is completed in order to benchmark the performance of three popular deep learning frameworks over the AI computing device. Experimental results demonstrate that some deep learning frameworks utilize the resources of GPU power while others utilize the multicore ARM-CPU system of the device for its parallel model of computation. Fourth, the robustness of state-of-the-art pre-trained deep learning models for feature extraction is analyzed and contrasted with the previous approach used by PABI. Experimental results indicate that pre-trained deep learning models overcome the traditional approaches in some fields; however, combining different pre-trained models in a process reduces its accuracy. Fifth, a patient-tracking algorithm based on an identity verification approach is developed to improve the autonomy, usability, and interactions of patients with the robot. Experimental results show that the developed algorithm has the potential to perform as well as the previous algorithm used by PABI based on a deep learning classifier approach.

In the field of Cybersecurity, it is essential to know who is connected to your system. The functionality for Authentication of connecting users in the local area network is in the focus for this report. There exist various authentication protocols, however in this report IEEE 802.1X is covered since it is the protocol most suitable for wired local area networks. The IEEE 802.1X protocol is studied in theory with its architecture of Supplicator, Authenticator and Authentication server and the used communication protocols EAPOL and RADIUS. A practical test was then performed as a basic concept to learn more about pros and cons for utilizing these protocols where the fundamentals of protocol communications are observed and later the prerequisites for a larger scale implementation are described.  The outcome from the test is proof of the relative difficulties involved with having to keep up with the pace of Cybersecurity evolution. In the test, older equipment where thought to be used, however due to incompatibility of gear and software the test needed to be revised to use other gear. The learning outcome from the test is that it is a complex task to set up authentication, competent staff are needed, as well as suitable equipment.  The motivation for setting up IEEE 802.1X is found in larger organizations where the risks of an attack are high, the large number of users calls for centralized systems for the handling of users and network policies. Due to the trend of Bringing you own device, a policy for the handling of unauthorized users and devices is needed to be in place. The default behavior may be to just deny access for unauthorized devices, however with authentication systems implemented the unauthorized user may instead benefit from being automatically referred to a guest network in a secured manner and the authorized user gains flexibility to access the network thru any available network port.  For the improvement and maintenance of Cybersecurity administration an Information Security Management System is found useful, the organization can thereby continuously improve their work and document the system features and routines. In case of a security breach that system gives support for immediate action upon the problem, and even stronger preparation for the Cyber defense in the form of good backup routines and monitoring the normal state activities where all devices are either authorized or unauthorized and placed into their proper network according to network policies. / <p>Examensarbete för högskoleingenjörsexamen i nätverksteknik</p>

Examensarbetets mål är att bygga en Quadcopter som kan flyga och styras via en radio-sändare i alla riktningar, stabilisera sig själv i luften, kunna landa autonomt och den skall även motverka krock i framåtgående riktning.För att uppnå grundmålet att kunna flyga så implementerades en PID-kontroller som används för att stabilisera Quadcoptern i luften genom att reglera motorer efter sensor-orientering. Denna sensororientering fås kombinera vinkeldata från en accelerometer och ett gyroskop.För att uppnå målet med autonom landning så användes en ultraljudssensor. En egen algoritm utvecklades för att läsa avstånd från marken. Avståndet från marken användes som grund för att skapa en egen algoritm för den autonoma landningen.Quadcoptern kan i slutprodukten stabilisera sig själv i luften, styras via radio och landa autonomt. Säkerhetsfunktioner som att aktivera autonom landning ifall Quadcoptern kommer utanför radions täckning är implementerade och även en brytare som stänger av motorerna. Det enda som inte implementerades var krocken i framåtgående riktning eftersom att en ultraljudssensor inte var lämpad för detta användningsområde. / The goal of the thesis is to build a quadcopter that can fly and be controlled via a radio transmitter in all directions, stabilizing itself in the air, to land autonomously and detect collisions in the forward direction.In order to achieve the basic goal of being able to fly, a PID controller was implemented which is used to stabilize the Quadcopter in the air by controlling the motors with help from sensor orientation. This sensor orientation is obtained from a complementary filter that merges angle data from an accelerometer and a gyroscope. Both an accelerometer and a gyroscope are required to automatically stabilize the Quadcopter in the air.To achieve the goal of autonomous landing an ultrasonic sensor was used. An algorithm was developed to read the distance from the ground which was a basis for creating our own algorithm for autonomous landing.The Quadcopter has the ability to stabilize itself in the air, be controlled via a radio transmitter and land autonomously. A safety feature that enables autonomous landing if the Quadcopter travels outside the radio coverage is implemented and also a switch that turns off the engines. The only thing that wasn’t implemented was the crash avoidance in the forward direction because the ultrasonic sensor was not suited for this application.

Deploying an embedded system to act as a controller for electronics is not new.  Today these kinds of systems are all around us and are used for a multitude of purposes. In contrast, cloud computing is a relatively new approach for computing as a whole. This thesis project explores these two technologies in order to create a bridge between these two wildly different platforms. Such a bridge should enable new ways of exposing features and doing maintenance on embedded devices. This could save companies not only time and money while dealing with maintenance tasks for embedded systems, but this should also avoid the needed to host this maintenance software on dedicated servers – rather these tasks could use cloud resources only when needed. This thesis explores such a bridge and presents techniques suitable for joining these two computing paradigms together. Exploring what is included in cloud computing by examining available technologies for deployment is important to be able to get a picture of what the market has to offer. More importantly is how such a deployment can be done and what the benefits are. How technologies such as databases, load-balancers, and computing environments have been adapted to a cloud environment and what draw-backs and new features are available in this environment are of interest and how a solution can exploit these features in a real-world scenario.  Three different cloud providers and their products have been presented in order to create an overview of the current offerings.  In order to realize a solution a way of communicating and exchanging data is presented and discussed. Again to realize the concept in a real-world scenario. This thesis presents the concept of cloud connectivity for embedded systems. Following this the thesis describes a prototype of how such a solution could be realized and utilized. The thesis evaluates current cloud providers in terms of the requirements of the prototype. A middle-ware solution drawing strengths from the services offered by cloud vendors for deployment at a vendor is proposed. This middle-ware acts in a stateless manner to provide communication and bridging of functionality between two parties with different capabilities. This approach creates a flexible common ground for end-user clients and reduces the burden of having the embedded systems themselves process and distribute information to the clients.  The solution also provides and abstraction of the embedded systems further securing the communication with the systems by it only being enabled for valid middle-ware services. / Att använda ett inbyggt system som en kontrollenhet för elektronik är inget nytt. Dessa typer av system finns idag överallt och används i vidt spridda användningsområden medans datormolnet är en ny approach för dator användning i sin helhet. Utforska och skapa en länk mellan dessa två mycket olika platformar för att facilitera nya tillvägagångs sätt att sköta underhåll sparar företag inte tid och pengar när det kommer till inbyggda system utan också när det gäller driften för servrar. Denna examensarbete utforskar denna typ av länk och presenterar för endamålet lämpliga tekniker att koppla dem samman medans lämpligheten för en sådan lösning diskuteras. Att utforska det som inkluderas i konceptet molnet genom att undersöka tillgängliga teknologier för utveckling är viktigt för att få en bild av vad marknaden har att erbjuda. Mer viktigt är hur utveckling går till och vilka fördelarna är. Hur teknologoier som databaser, last distrubutörer och server miljöer har adapterats till molnmiljön och vilka nackdelar och fördelar som kommit ut av detta är av intresse och vidare hur en lösning kan använda sig av dessa fördelar i ett verkliget scenario. Tre olika moln leverantörer och deras produkter har presenterats för att ge en bild av vad som för tillfället erbjuds. För att realisera en lösning har ett sett att kommunicera och utbyta data presenterats och diskuterats. Åter igen för att realisera konceptet i ett verkligt scenario. Denna uppsats presenterar konceptet moln anslutbarhet för inbyggda system för att kunna få en lösning realiserad och använd. En mellanprograms lösning som drar styrka ifrån de tjänster som erbjudas av molnleverantörer för driftsättning hos en leverantor föreslås. Denna mellanprogramslösnings agerar tillståndslöst för att erbjuda kommunikation och funktions sammankoppling mellan de två olika deltagarna som har olika förutsätningar. Denna approach skapar en flexibel gemensam plattform för olika klienter hos slutanvändaren och minskar bördan hos de inbyggdasystemet att behöva göra analyser och distrubuera informationen till klienterna. Denna lösning erbjuder också en abstraktion av de inbyggdasystemen för att erbjuda ytterligare säkerhet när kommunikation sker med de inbyggdasystemet genom att den endast sker med giltiga mellanprogram.

As space missions continue to increase in complexity, the operational capabilities and amount of gathered data demand ever more advanced systems. Currently, mission capabilities are often constrained by the link bandwidth as well as onboard processing capabilities. A large number of commands and complex ground station systems are required to allow spacecraft operations. Thus, methods to allow more efficient use of the bandwidth, computing capacity and increased autonomous capabilities are of strong research interest. Artificial Intelligence (AI), with its vast areas of application scenarios, allows for these challenges and more to be tackled in the spacecraft design. Particularly, the flexibility of Artificial Neural Networks as Machine Learning technology provides many possibilities. For example, Artificial Neural Networks can be used for object detection and classification tasks. Unfortunately, the execution of current Machine Learning algorithms consumes a large amount of power and memory resources. Additionally, the qualification of such algorithms remains challenging, which limits their possible applications in space systems. Thus, an increase in efficiency in all aspects is required to further enable these technologies for space applications. The optimisation of the algorithm for System on Chip (SoC) platforms allows it to benefit from the best of a generic processor and hardware acceleration. This increased complexity of the processing system shall allow broader and more flexible applications of these technologies with a minimum increase of power consumption. As Commercial off-the-shelf embedded systems are commonly used in NewSpace applications and such SoC are not yet available in a qualified manner, the deployment of Machine Learning algorithms on such devices has been evaluated. For deployment of machine learning on such devices, a ConvolutionalNeural Network model was optimised on a workstation. Then, the neural network is deployed with Xilinx’s Vitis AI onto a SoC which includes a powerful generic processor as well as the hardware programming capabilities of an Field ProgrammableGate Array (FPGA). This result was evaluated based on relevant performance and efficiency parameters and a summary is given in this thesis. Additionally, a tool utilising a different approach was developed. With a high-level synthesis tool the hardware description language of an accelerated linear algebra optimised network is created and directly deployed into FPGA logic. The implementation of this tool was started, and the proof of concept is presented. Furthermore, existing challenges with the auto-generated code are outlined and future steps to automate and improve the entire workflow are presented. As both workflows are very different and thus aim for different usage scenarios, both workflows are outlined and the benefits and disadvantages of both are outlined.

The IEEE Power over Ethernet (PoE) protocol is currently the most used standard for allowing Power Sourcing Equipment (PSE) to distribute power to Powered Devices (PD) through standard Ethernet cables, permitting devices to be added to a network without needing extra power cables.There exist many PSE manager circuits which can be used to easily distribute power to a handful of PDs, but a problem arises when more are desired, since very few of these circuits can communicate with other circuits, and thus can't coordinate power distribution with each other.The goal of this work was to discover the feature set of some popular PSE managers, discover any real-time requirements they may demand of a host, and to then use this knowledge to create an API which could coordinate multiple PSE managers to handle power distribution.The work was carried out with a study of three PSE managers (PD69200, LTC4291, and TPS23880), as well as study and experimentation of the real-time requirements. After this a series of different architectures were created and evaluated using a combination of analysis, questionnaires, and prototype implementation. The work resulted in a software architecture which allows any number of PSE managers of any model to be used simultaneously in a system, with minimal-to-zero considerations needing to be made for different PSE managers. This allows the system hardware to be designed free of any software considerations, and allows the user to operate the system similarly regardless of the underlying architecture.

Caesarcipher is a commonly known encryption method although it is known for being unsure, even long before computers were invented. Even though it is unsure it can still be used for cryptography, JackTheRidder consists of two main algorithms where one of them is based on the Caesarcipher. With many rotations of Caesarciphers on different parts of texts, like many Vigenerecryptos of different sizes overlapping each other, makes the encryption secure. The other main algorithm of JackTheRidder is based on the Collatz Conejcture, which is a famous unsolved mathematical problem. This algorithm counts the amount of steps it takes before the chosen number reaches 1 and then moves the letter with those steps which results in the encryption. The amount of steps is a result from a positive integer going through Collatz Conjecture, which divides the integer if its even and uses (3𝑥+1)/2 if the integer is odd. The algorithm then moves to the next character to encrypt and adds one to the integer used. JackTheRidder is an encryption method created by the authors of this project and further will be analyzed in this essay. The focus of this essay is to analyze how Collatz Conjecture and Caesarcipher functions in encryption methods. JackTheRidder is going to be examined through literature study and experiments. There are three criterias that JackTheRidder is going to be analyzed from; security, robustness and efficiency. The results indicate that JackTheRidder needs to be optimized and be written in machine code for it to be able to get better implemented in society. JackTheRidder is proven to be better than AES in the chi2 test which is an indicator that it might be a viable option for replacement. Overall in the experiments, JackTheRidder held a high level which is a positive outcome but still it is not guaranteed to be secure enough to be implemented into the society at this stage of developement.

With increasing popularity of indoor farming, more and more home growers are faced with sub-optimal lighting conditions in northern countries or poorly lit windows. We have designed and built a proof-of-concept system capable of reducing electricity cost and CO2 footprint of the electricity used for consumer grade grow lights without adversely impacting the grow cycle of the plant. Our system provides optimal grow light conditions for a given plant while using forecasts and live grid data from the ENTSO-E transparency platform to automatically use or store electricity during low-cost hours and avoid using grid electricity during high-cost hours, but can also be configured to prioritize electricity use when the available grid power’s carbon intensity is low. The system, consisting of a server and an embedded control unit, was designed and implemented according to Nunamaker and Chen’s five-step iterative systems development research method and later evaluated by simulating the system for 14 days using real world sunlight and grid data. The results of the simulation show a significant reduction in both spending and carbon emissions related to electricity use, with figures of 73% and 28%, respectively. However, when accounting for life-cycle cost and emissions from the battery, the prototype in its current configuration is neither profitable nor a net positive for the environment. With changes to battery type and taking advantage of economies of scale, a future version could be economically viable, but to be environmentally sustainable, further advances in eco-friendly battery production are needed.

In the recent years, there has been a rapid development of miniature sensor based wearable devices with broad functionality. Some of these devices are designed to function continuously for weeks with a single battery charge, which requires their power usage to be as low as possible. This can be achieved either through circuit optimization, a more powerful battery, or through optimization of the software controlling the device. This thesis presents an approach to the latter option by the means of a power and energy simulator based on the Open Virtual Platform (OVP). Such a tool can simulate the embedded software and predict the power and energy usage of the system. The result of the simulator could be used in combination with other tools to optimize the software implementation of the system and to lower its overall power and energy consumption.