ITC/USA 2010 Conference Proceedings / The Forty-Sixth Annual International Telemetering Conference and Technical Exhibition / October 25-28, 2010 / Town and Country Resort & Convention Center, San Diego, California / When designing data transport systems, Telemetry and Communications engineers always face the risk that their chosen hardware will not be available or supported soon after the hardware has been installed. The best way to reduce this risk and ensure the longevity of the system is to select an open architecture standard that is supported by multiple manufacturers. This open architecture should also have the ability to be easily upgraded and provide for all of the features and flexibility that are required to be a reliable carrier-grade edge-device. The PCI Industrial Computer Manufacturers Group (PICMG) developed the MicroTCA open standard to address the specific needs of these Communications and Network System Engineers. This paper describes the MicroTCA architecture and how it can be applied as the ideal edge-device solution for Networked Telemetry Systems applications.

The field of autonomous robots have been explored more and more over the last decade. The combination of machine learning advances and increases in computational power have created possibilities to explore the usage of machine learning models on edge devices. The usage of object detection on edge devices is bottlenecked by the edge devices' limited computational power and they therefore have constraints when compared to the usage of machine learning models on other devices. This project explored the possibility to use real time object detection on a Raspberry Pi as input in different control systems. The Raspberry with the help of a coral USB accelerator was able to find a specified object and drive to it, and it did so successfully with all the control systems tested. As the robot was able to navigate to the specified object with all control systems, the possibility of using real time object detection in faster paced situations can be explored. / Ämnet autonoma robotar har blivit mer och mer undersökt under de senaste årtiondet. Kombinationen av maskin inlärnings förbättringar och ökade beräknings möjligheter hos datorer och chip har gjort det möjligt att undersöka användningen av maskin inlärningsmodeller på edge enheter. Användandet av bildigenkänning på edge enheter är begränsad av edge enheten begränsade datorkraft, och har därför mer begränsningar i jämförelse med om man använder bildigenkänning på en annan typ av enhet. Det här projektet har undersökt möjligheten att använda bildigenkänning i realtid som input för kontrollsystem på en Raspberry Pi. Raspberry Pien med hjälp av en Coral USB accelerator lyckades att lokalisera och köra till ett specificerat objekt, Raspberryn gjorde detta med alla kontrollsystem som testades på den. Eftersom roboten lyckades med detta, så öppnas möjligheten att använda bildigenkänning på edge enheter i snabbare situationer.

This degree project proposes a novel architecture for IoT systems, utilizing smartphones as edge devices and running the value-creating software such as preprocessing, anomaly detection, and deriving data-based insights in the web browser as opposed to natively on the device. Utilizing the smartphone as an edge device reduces cost of adoption for IoT technologies since less hardware has to be included in the system compared to bundling a device for processing with the system. However, in typical implementations, these smartphones are running native applications that are necessarily bound by the policy set up by the owners of the major application marketplaces through which the applications are distributed, that among other things enforce a heavy revenue-sharing program. Furthermore, this sytem is not convenient for the user since they have to download an application they may only use once or sparingly, and it relies on the compatibility between the operating system and the application. Running the application in the browser as opposed to natively would solve these issues. This work also explores the experimental WebUSB API and investigates its applicability in IoT systems. Specifically, this work explores two constructed scenarios that showcase the promise of this novel architecture to extrapolate how it can be utilized for other purposes. These two constructed scenarios are successfully implemented, and various metrics to analyse their performance and real world applicability are discussed. In one experiment, a QR-scanning application is implemented in the browser, and showcases an average frame rate of over 60 frames per second while rendering a live video feed of the contents captured by the camera as well as a loading animation, and an average time to completion for scanning a QR code of 0,204 seconds after initiating the scan. In another experiment, a firmware update is simulated by transferring encrypted data from the browser via the WebUSB API to a microcontroller. Due to the limitations of the experimental setup, the implementation could showcase a transfer of no more than 29 KB of encrypted data in 10 seconds. However, the implementation successfully shows that the browser can remain interactive even while performing these transferring operations, and that there are good APIs in place for developers to easily access the advanced sensors of the phone, and that the WebUSB API has good safety measures in place. Furthermore, the work successfully demonstrates how the WebUSB API can be utilized in IoT systems as a novel way of transferring data that holds great implications for the future of IoT systems in general, and the web in particular. To conclude, the work finds that the modern web browser works well as an environment for IoT applications, and that it has good access to the advanced sensors of the smartphone, and that theWebUSB API can effectively be utilised for data transfer in IoT applications. / Detta examensarbete föreslår en ny arkitektur för IoT-system som använder smarta telefoner som edge-enheter som behandlar den data de tar emot i webbläsaren, snarare än nativt på enheten. Att använda smarta telefoner som en edge-enheter minskar kostnaden för användning av IoT-teknologier eftersom mindre hårdvara behöver ingå i systemet, eftersom viss funktionalitet istället kan substitueras av den smarta telefonen. I typiska implementationer kör dessa smarta telefoner nativa applikationer som nödvändigtvis är bundna av de regelverk som fastställts av ägarna av de stora applikationsmarknaderna genom vilka applikationerna distribueras, vilket bland annat påtvingar applikationsägarna att ge bort stora delar av sina intäkter. Dessutom är detta system inte bekvämt för användaren eftersom de måste ladda ner en applikation de bara kommer använda ett fåtal gånger, och som dessutom är beroende av kompatibiliteten med operativsystemet. Att istället köra en motsvarande applikation i webbläsaren löser dessa problem. Detta arbete utforskar också det experimentella WebUSB API:t och undersöker dess användbarhet i IoT-system. Specifikt utforskar detta arbete två konstruerade scenarier som visar potentialen i denna nya arkitektur för att extrapolera hur den kan användas för andra ändamål. Dessa två konstruerade scenarier implementeras framgångsrikt och olika mått för att analysera deras prestanda och verkliga tillämpbarhet diskuteras. I ett experiment implementeras en QR- skanningsapplikation i webbläsaren och uppnår en genomsnittlig bildfrekvens på över 60 bildrutor per sekund samtidigt som en live videoström av innehållet som fångas av kameran och en laddningsanimation visas på skärmen. Vidare uppnådde applikationen en genomsnittlig tid för tolkning av QR-koder på 0,204 sekunder, från och med det att skanningen inleddes. I ett annat experiment simuleras en programvaruuppdatering genom att överföra krypterad data från webbläsaren via WebUSB API:t till en enkretsdator. På grund av begränsningarna i experimentet kunde implementationen inte visa en överföring högre än 29 KB krypterad data på 10 sekunder. Implementationen visar dock framgångsrikt att webbläsaren kan förbli interaktiv även när de utför dessa överföringsåtgärder, och att det finns bra API: er för utvecklare att enkelt få tillgång till telefonens avancerade sensorer och att WebUSB API:t har goda säkerhetsmekanismer på plats. Dessutom demonstrerar arbetet framgångsrikt hur WebUSB API:t kan användas i IoT-system som ett nytt sätt att överföra data som har stora implikationer för framtiden för IoT-system i allmänhet och på webbläsare och webbapplikationer synnerhet. Avslutningsvis konstaterar arbetet att den moderna webbläsaren fungerar bra som en miljö för IoT-system och att den har god tillgång till den smarta telefonens avancerade sensorer och att WebUSB API:t framgångsrikt kan användas för dataöverföring i IoT-system.

Edge computing is the idea of moving computations away from the cloud andinstead perform them at the edge of the network. The benefits of edge computing arereduced latency, increased integrity, and less strain on networks. Edge AI is the practiceof deploying machine learning algorithms to perform computations on the edge.In this project, a pre-trained model YAMNet is retrained and used to perform audioclassification in real-time to detect gunshots, glass shattering, and speech. The modelis deployed onto the edge device both as a full TensorFlow model and as TensorFlowLite models. Comparing results of accuracy, inference time, and memory allocationfor full TensorFlow and TensorFlow Lite models with and without optimization. Resultsfrom this research were that it was a valid option to use both TensorFlow andTensorFlow Lite but there was a lot of performance to gain by using TensorFlow Litewith little downside.

Interconnected devices are becoming attractive solutions to integrate physical parameters and making them more accessible for further analysis. Edge devices, located at the end of the physical world, measure and transfer data to the remote server using either wired or wireless communication. The exploding number of sensors, being used in the Internet of Things (IoT), medical fields, or industry, are demanding huge bandwidth and computational capabilities in the cloud, to be processed by Artificial Neural Networks (ANNs) – especially, processing audio, video and images from hundreds of edge devices. Additionally, continuous transmission of information to the remote server not only hampers privacy but also increases latency and takes more power. Deep Neural Network (DNN) is proving to be very effective for cognitive tasks, such as speech recognition, object detection, etc., and attracting researchers to apply it in edge devices. Microcontrollers and single-board computers are the most commonly used types of edge devices. These have gone through significant advancements over the years and capable of performing more sophisticated computations, making it a reasonable choice to implement DNN. In this thesis, a DNN model is trained and implemented for Keyword Spotting (KWS) on two types of edge devices: a bare-metal embedded device (microcontroller) and a robot car. The unnecessary components and noise of audio samples are removed, and speech features are extracted using Mel-Frequency Cepstral Co-efficient (MFCC). In the bare-metal microcontroller platform, these features are efficiently extracted using Digital Signal Processing (DSP) library, which makes the calculation much faster. A Depth wise Separable Convolutional Neural Network (DSCNN) based model is proposed and trained with an accuracy of about 91% with only 721 thousand trainable parameters. After implementing the DNN on the microcontroller, the converted model takes only 11.52 Kbyte (2.16%) RAM and 169.63 Kbyte (8.48%) Flash of the test device. It needs to perform 287,673 Multiply-and-Accumulate (MACC) operations and takes about 7ms to execute the model. This trained model is also implemented on the robot car, Jetbot, and designed a voice-controlled robotic vehicle. This robot accepts few selected voice commands-such as “go”, “stop”, etc. and executes accordingly with reasonable accuracy. The Jetbot takes about 15ms to execute the KWS. Thus, this study demonstrates the implementation of Neural Network based KWS on two different types of edge devices: a bare-metal embedded device without any Operating System (OS) and a robot car running on embedded Linux OS. It also shows the feasibility of bare-metal offline KWS implementation for autonomous systems, particularly autonomous vehicles.<br>

Antennas are one of the most sensitive elements in any wireless communication equipment. Designing small-profile, multiband and wideband internal antennas with a simple structure has become a necessary challenge. In this thesis, two planar antennas are designed, simulated and implemented on an effort to cover the LTE-M1 and NB-IoT radio frequencies. The cellular antenna is designed to receive and transmit data over the eight-band LTE700/GSM/UMTS, and the GNSS antenna is designed to receive signal from the global positioning system and global navigation systems, GPS (USA) and GLONASS. The antennas are suitable for direct print on the system circuit board of a device. Related theory and research work are discussed and referenced, providing a strong configuration for future use. Recommendations and suggestions on future work are also discussed. The proposed antenna system is more than promising and with further adjustments and refinement can lead to a fully working solution.

In recent years, AI has developed significantly and become more popular than ever before. The applications of AI are expanding, making knowledge about its application and the systems it can be applied to more important. This project compares and evaluates deep learning models for object detection of traffic signs on the embedded systems Nvidia Jetson Nano and Raspberry Pi 3 Model B. The project compares and evaluates the models YOLOv5, SSD Mobilenet V1, FOMO, and Efficientdet-lite0. The project evaluates the performance of these models on the aforementioned embedded systems, measuring metrics such as CPU usage, FPS and RAM. Deep learning models are resource-intensive, and embedded systems have limited resources. Embedded systems often have different types of processor architectures than regular computers, which means that some frameworks and libraries may not be compatible. The results show that the tested systems are capable of object detection but with varying performance. Jetson Nano performs at a level we consider sufficiently high for use in production depending on the specific requirements. Raspberry Pi 3 performs at a level that may not be acceptable for real-time recognition of traffic signs. We see the greatest potential for Efficientdet-lite0 and YOLOv5 in recognizing traffic signs. The distance at which the models detect signs seems to be important for how many signs they find. For this reason, SSD MobileNet V1 is not recommended without further trai-ning despite its superior speed. YOLOv5 stood out as the model that detected signs at the longest distance and made the most detections overall. When considering all the results, we believe that Efficientdet-lite0 is the model that performs the best. / Under de senaste åren har AI utvecklats mycket och blivit mer populärt än någonsin. Tillämpningsområdena för AI ökar och därmed blir kunskap om hur det kan tillämpas och på vilka system viktigare. I det här projektet jämförs och utvärderas djupinlärningsmodeller för objektdetektering av trafikskyltar på de inbyggda systemen Nvidia Jetson Nano och Raspberry Pi 3 Model B. Modellerna som jämförs och utvärderas är YOLOv5, SSD Mobilenet V1, FOMO och Efficientdet-lite0. För varje modell mäts blandannat CPU-användning, FPS och RAM. Modeller för djupinlärning är resurskrävande och inbyggda system har begränsat med resurser. Inbyggda system har ofta andra typer av processorarkitekturer än en vanlig dator vilket gör att olika ramverk och andra bibliotek inte är kompatibla. Resultaten visar att de testade systemen klarar av objektdetektering med varierande prestation. Jetson Nano presterar på en nivå vi anser vara tillräckligt hög för användning i produktion beroende på hur hårda krav som ställs. Raspberry Pi 3 presterar på en nivå som möjligtvis inte är acceptabel för igenkänning av trafikskyltar i realtid. Vi ser störst potential för Efficientdet-lite0 och YOLOv5 för igenkänning av trafikskyltar. Hur långt avstånd modellerna upptäcker skyltar på verkar vara viktigt för hur många skyltar de hittar. Av den anledningen är SSD MobileNet V1 inte att rekommendera utan vidare träning trots sin överlägsna hastighet. YOLOv5 utmärkte sig som den som upptäckte skyltar på längst avstånd och som gjorde flest upptäckter totalt. När alla resultat vägs in anser vi dock att Efficientdet-lite0 är den modell som presterar bäst.