Protecting personal privacy has become an increasingly important issue as computers become a more integral part of everyday life. As people begin to trust more personal information to be contained in computers they will question if that information is safe from unwanted intrusion and access. With the rise of mobile devices (e.g., smartphones, tablets, wearable technology) users have enjoyed the convenience and availability of stored personal information in mobile devices, both in the operating system and within applications.For a mobile application to function correctly it needs permission or privileges to access and control various resources and controls on the mobile device. These permissions can range from location and account information to access to all storage on the mobile device. A single permission, or a combination of permissions, could lead to a high risk of potential privacy invasion. This privacy invasion risk can be amplified specifically for security applications when compared to non-security applications due to the administrative privileges that security applications frequently need to moderate and protect information on a mobile device. Currently there is no defined matrix or framework for analyzing privacy risks for any mobile platform, including the main mobile platforms of Android, iOS and Windows mobile.The purpose of this research is to create a framework for analyzing mobile application permissions and identify potentially invading permission. The framework produces a Privacy Invasion Profile (also known as a PIP) for each application, which can be used to compare the risk of privacy invasion for a specific application.

<p>Pathogen detection via viscosity quantification in biological systems has long been an essential aspect of biomedical research. The importance of persistent testing of pathogens such as <em>V. cholerae</em> and HIV has consistently been recognized but limited in regions where systematic and financial resources are unavailable. Current methods require the samples be transported to research labs primarily in large cities or different countries. For consistent pathogen testing to be performed in remote areas, detection methods must be designed for portability with laboratory standards and simplicity for use without much technical background in place. </p> <p>Particle Diffusometry is a visualization method on the result of the amplification of pathogen by quantifying the Brownian motion of suspended particles in a solution. The amplification usually occurs in the specialized machine; then, the fluid sample gets inserted into the microfluidic chip for optical observation for Brownian motion. The technique has been used in particle sizing and measuring viscosity change in the biomolecular solution. In use with limitations, I present the improvements on the existing Particle Diffusometry technique to expand its use in broader biomedical applications.</p> <p>We address the portability of the technique. In the emerging and fast-growing mobile technology market, we have developed a smartphone-based portable platform capable of performing par quality tasks compared to traditional lab-based microscopy. We successfully measure the presence of <em>V. cholerae</em> as few as 6 cells/reaction, a waterborne pathogen, where its DNA is spiked into environmental water sources in just under 35 minutes. To further make the overall technology portable, we developed an on-chip amplification method accompanied by the portable heating unit. A mobile heating unit removes the need for the qPCR machine to amplify the biomolecular structure. Also, it opens the capability of on-chip amplification, further simplifying the steps needed to identify the pathogen in the source. We confirm the validity of the developed setup by measuring the presence of as low as 50 SARS-CoV-2 virus particles within 10% saliva. </p> <p>Addressing two main limitations of the existing Particle Diffusometry technique, improvements in the algorithm occur. First, we improve the algorithm to calculate diffusion coefficients even when the particles suspended in the sample are experiencing unified patterns, hence the flow, when recording is taking place. The improved algorithm correctly identified the diffusion coefficient within  margin of error using simulation and experimental verification for the sample under simple shear flow types, uniform, Couette, and Poiseuille. Second, we address the mismatch between the frame rate of the camera and the Brownian motion of particles at elevated temperatures. By configuring the correction equation for the frame mismatch behavior, we corrected the deviation of the diffusion coefficient in the range of 3E-13 to 3E-12 m²/s. Ultimately, we applied the improved flow algorithm to the elevated temperature simulation, showing the error propagation does not differ by the temperature; the percentage of error in computing the diffusion coefficient for the sample exhibiting flow only depends on the flow velocity. </p> <p>Applying these two improvements, we perform measurements on over-time viscosity change using the hydrogel formation. We characterize the hydrogel formation time using the diffusion gradient plane and variation of the initiator. By applying the addressed improvements on the real-time detection of HIV amplification on-chip, we further validate the applicative nature of the extended Particle Diffusometry technique. </p> <p>Real-time flow-adjusted Particle Diffusometry is, therefore, a feasible method for detecting viscosity changes in both chemical and biomolecular solutions in real-time. This approach opens up an alternative method for measuring biological amplification in real-time. The improvements further open the existing Particle Diffusometry technique to be widely used in the field involving rheology and pathogen detection not only in the traditional lab-based setting but also out in the field. </p>

Mobiltelefoner, sociala medier och smarta teknologier är idag en självklar del av tillvaron för många. Samtidigt har överanvändning av mobiltelefoner förknippats med psykisk ohälsa. Som del av utvecklingen av smarta telefoner, appar och tillgången till stora datamängder har även fenomenet att bevaka sina egna vanor med hjälp av tekniken blivit vanligt förekommande. Denna typ av informationsinsamling kallas för quantified self-data och intresset för tekniken är stort. Tidigare forskning om quantified self-data antyder att det skulle kunna vara ett kraftfullt hjälpmedel för en användare som vill förbättra sin livskvalitet. Ett däremot outforskat område är vad som händer med effekterna av quantified self när det ställs mot teknologier som redan vunnit mark i att styra en användares beteende, exempelvis våra smartphones. Denna studie syftade till att utforska detta område och frågan som ställs är hur kontinuerlig feedback i form av quantified self-data, via en mobilapplikation, har för inverkan på en mobilanvändares upplevelse av mobilen och sin användning av den. I studien fick åtta deltagare under två veckor använda sig av mobilapplikationen Moment, som presenterar data om en användares mobilanvändning. Deltagarna ombads även skriva ner dagliga reflektioner i en loggbok. Resultatet visar på att appen tillsammans med loggboken fick deltagarna att aktivt reflektera över sin användning. Deltagarna ansåg att deras användning förblev oförändrad, men att appen och dess data mestadels motiverade till ett minskat användande. Samtliga deltagare trodde att appen var ett bra hjälpmedel för den som vill minska sin mobilanvändning. / Smartphones and social media have become common parts of our everyday life. At the same time, smartphone attachment and overuse have been associated with negative implications on mental health. As a part of the development of smartphones, apps and the access to great amounts of data, the phenomenon of tracking one’s own habits is frequently occurring. This type of data gathering is called quantified self and there is a big interest in the technology. Previous research suggests that quantified self could be a powerful tool for a user wishing to improve his or her life quality. However, the effect of quantified self when put up against technology that is already effective in steering users’ behaviors, such as our smartphones, remains an unexplored subject. This study aimed to explore that subject and the questioned asked is how continuous feedback through quantified self, when presented in a mobile application, influences a smartphone user’s experience of the phone and the usage of it. In the study conducted, eight participants used the application Moment during two weeks, an application which presents the user with data about phone usage. Participants where prompted to write down daily reflections in a journal. The result shows that the app together with the journal made the participants actively reflect on their phone usage. Participants felt that their usage remained unchanged, but that the app mostly motivated them to reduce their usage. All participants thought that the app would be a good aid for those who want to reduce their smartphone usage.

Målet med detta examensarbete har varit att utveckla en app med en algoritm för att mäta rotationshastigheten hos en prototyp för en spirometerlösning till en låg kostnad för smart- telefoner. En förstudie har gjorts av smarttelefoners användbarhet för att mäta hälsotillstånd och vilka alternativa lösningar och algoritmer som finns för att mäta rörelsedetektion. I detta arbete har en app med en algoritm utvecklats för att detektera rörelser och mäta rotations- hastigheten hos spirometerprototypens turbin filmad med en smarttelefonkamera. För att metoden ska fungera är det viktigt att rotationshastigheten understiger hälften av kamerans fps (bilder per sekund). Rotationshastigheten hos turbinen måste därför begränsas och det behövs en kamera som klarar minst 120 fps för att fånga rörelserna i prototypens turbin.Arbetet har resulterat i en fungerande algoritm för att bestämma turbinens rotationshastighet. Den utvecklade algoritmen detekterar topparna i en PPG (photoplethysmogram). För att minska beräkningstiden och för att öka noggrannheten analyserar algoritmen färgintensiteten i ett begränsat område, ett s.k. ROI (Region of Interest) i varje bild. Det finns stora möjligheter att använda denna algoritm för att fortsätta utvecklingen av detta alternativa sätt att utföra spirometritester. / The goal with this bachelor thesis was to develop an application with an algorithm to measure the rotation speed of a prototype, as a low-cost solution for measuring spirometry with a smartphone. In a pilot study it was investigated how a smartphone can be used to measure health and what algorithms there are to detect motion in videos. After the pilot study an app with the function to record a video by using the camera of a smartphone and then use an algorithm to detect the rotation speed in the spirometry-prototype’s turbine was developed. To make it work it is important that the rotation speed is low enough so it does not exceed half of the cameras fps. Therefore, to capture the rotation speed of the spirometry-prototype’s turbine the rotation needs to be limited and a smartphone with a camera with at least 120 fps is required.The result of this work is an algorithm that can measure the rotation speed in the spirometry prototype turbine. The algorithm is detecting the peaks in a PPG. To minimize the computation time and to increase the accuracy the algorithm analyses the colour intensity over a ROI in every frame. There is great potential to use this algorithm to further develop this alternative method of measuring spirometry.

In this report we attempt to provide insights to how a resource efficient solution for transportation mode recognition can be implemented on a smartphone using the accelerometer and magnetometer as sensors for data collection. The proposed system uses a hierarchical classification process where instances are first classified as vehicles or non-vehicles, then as wheel or rail vehicles, and lastly as belonging to one of the transportation modes: bus, car, motorcycle, subway, or train. A virtual gyroscope is implemented as a low-power source of simulated gyroscope data. Features are extracted from the accelerometer, magnetometer and virtual gyroscope readings that are sampled at 30 Hz, before they are classified using machine learning algorithms from the WEKA machine learning library. An Android application was developed to classify real-time data, and the resource consumption of the application was measured using the Trepn profiler application. The proposed system achieves an overall accuracy of 82.7% and a vehicular accuracy of 84.9% using a 5 second window with 75% overlap while having an average power consumption of 8.5 mW. / I denna rapport försöker vi ge insikter om hur en resurseffektiv lösning för transportlägesigenkänning kan implementeras på en smartphone genom att använda accelerometern och magnetometern som sensorer för datainsamling. Det föreslagna systemet använder en hierarkisk klassificeringsprocess där instanser först klassificeras som fordon eller icke-fordon, sedan som hjul- eller järnvägsfordon, och slutligen som tillhörande ett av transportsätten: buss, bil, motorcykel, tunnelbana eller tåg. Ett virtuellt gyroskop implementeras som en lågenergi källa till simulerad gyroskopdata. Olika särdrag extraheras från accelerometer, magnetometer och virtuella gyroskopläsningar som samlas in vid 30 Hz, innan de klassificeras med hjälp av maskininlärningsalgoritmer från WEKA-maskinlärningsbiblioteket. En Android-applikation har utvecklats för att klassificera realtidsdata, och programmets resursförbrukning mättes med hjälp av Trepn profiler-applikationen. Det föreslagna systemet uppnår en övergripande noggrannhet av 82.7% och en fordonsnoggrannhet av 84.9% genom att använda ett 5 sekunders fönster med 75% överlappning med en genomsnittlig energiförbrukning av 8.5 mW.

It takes only about 10-30 minutes into a sedentary lecture before audience attention is decreasing. There are different ways to avoid this. One is to use a web-based audience response systems (ARS), where the audience interact with the lecturer through their smartphones, and another is to take short breaks, including physical movements, to re-energize both the body and the brain. In this study, these two methods have been combined and explored. By utilizing the motion sensors that are integrated in almost every smartphone, a physical activity for a lecture audience was created and implemented in the ARS platform Mentimeter. The proof of concept was evaluated in two lectures, based on O’Brien and Toms' model of engagement. The aim was to explore the prerequisites, both in terms of design and implementation, for creating an engaging physical activity within a lecture audience, using smartphone motion sensors to capture movements and a web-based ARS to present the data. The results showed that the proof of concept was perceived as fun and engaging, where important factors for creating engagement were found to be competition and a balanced level of task difficulty. The study showed that feedback is complicated when it comes to motion gesture interactions, and that there are limitations as to what can be done with smartphone motion sensors using web technologies. There is great potential for further research in how to design an energizing lecture activity using smartphones, as well as in exploring the area of feedback in motion gesture interaction. / Efter bara 10-30 minuter på en stillasittande föreläsning börjar publiken tappa i koncentration. Det går undvika på olika sätt. Ett sätt kan vara genom att låta publiken bli mer aktiva i föreläsningen med hjälp av ett webb-baserat röstningsverktyg, där de använder sina smartphones för att interagera med föreläsaren, och ett annat sätt kan vara att ta korta pauser där publiken får röra på sig för att syresätta hjärna och kropp. I den här studien kombinerades dessa två metoder genom att utnyttja rörelsesensorerna som finns inbyggda i de flesta smartphones. En fysisk aktivitet för en föreläsningspublik togs fram och implementerades i ARS-plattformen Mentimeter och konceptet utvärderades sedan under två föreläsningar baserat på O’Brien and Toms' modell för engagemang. Målet var att utforska förutsättningarna, både inom teknik och design, för att skapa en engagerande fysisk aktivitet för en föreläsningspublik, där smartphonens rörelsesensorer används för att fånga rörelse och ett webb-baserat röstningssystem för att presentera data. Resultatet visade att konceptet upplevdes som kul och engagerande, där viktiga faktorer för att skapa engagemang fanns i att ha ett tävlingsmoment och en lagom svårighetsgrad. Studien visade även att feedback är komplicerat när det kommer till rörelseinteraktion, och att det finns begräsningarna i vad som kan göras med rörelsesensorerna i en smartphone med hjälp av webbteknologi. Det finns stor potential för ytterligare undersökningar både inom hur man kan skapa interaktiva aktiviteter på föreläsningar som ger publiken mer energi, men också inom området kring feedback för rörelseinteraktion.

Widespread availability of digital tools has changed the daily lives of college students. Yet, the intricacies underlying these digital ecosystems and their relationship to psychological functioning, particularly among these younger ‘digital native’ age cohorts, are still unclear. A growing body of work points to associations between digital media behaviors and the capacity for top-down self-regulatory control over thoughts, emotions, and behavior. Behavioral scientists often subdivide this skillset into separate psychological constructs with different labels (e.g., cognitive control, executive functioning, emotion regulation), and use a varied array of tasks and surveys to index its subcomponents. The general finding from across behavioral studies is that groups (and individuals) with weaker executive functioning (EF) skills also tend to exhibit heavier and more problematic digital media habits (e.g., excessive, or addiction-like use). This is presumably because the inability to reliably exert control makes one more prone to impulsive engagement with digital media (e.g., frequent phone-checking), greater attentional distractibility in response to media-associated cues (e.g., notifications), and more difficulty with sustaining goal-relevant behaviors in the presence of digital media. However, there has also been empirical work suggesting null and even positive or nonlinear relationships between digital media use and EF. The current study aimed to address these seemingly opposing sets of findings by examining how, and to what extent, individual differences in one’s self- and mobile-reported smartphone habits relate with specific facets of higher-order cognition. In our examination of the interplay between these factors, we found consistent patterns emerge between subjective measures of everyday and problematic smartphone use and common non-social executive functioning skills. Furthermore, we also found evidence indicating an overlapping pattern of findings highlighting the relationships between one’s cognizance toward their actual mobile usage habits and specific facets of socially oriented self-regulation. / Psychology

This thesis addresses the problem of Road Damage Detection using object detection models,Yolov8 and Yolov5. While Yolov5 has been utilized in prior road damage detection projects, thiswork introduces the application of the newly released Yolov8 model to this domain. We haveprepared a dataset of 3,000 annotated images of road damage in Sweden and applied variousYolov8 and Yolov5 models to this dataset and a larger international one. The potential ofdeploying a lightweight Yolov8 model in a smartphone application for real-time detection, aswell as the effectiveness of an ensemble approach combining several models, were alsoexplored. The results show an F1 score of 0.57 and 0.6 for the best-performing models on theSwedish dataset and an international Road damage dataset respectively. Several box clusteringmethods were tested to combine the predictions of the ensemble, but none outperformed thebest individual model. A Quantized version of Yolov8 was deployed to a smartphone device withsatisfying performance. This work aims to create a model which can ultimately be used toimprove road safety and quality.T

Given concern over shark attacks in coastal regions, the recent use of unmanned aerial vehicles (UAVs), or drones, has increased to ensure the safety of beachgoers. However, much of city officials' process remains manual, with drone operation and review of footage still playing a significant role. In pursuit of a more automated solution, researchers have turned to the usage of neural networks to perform detection of sharks and other marine life. For on-device solutions, this has historically required assembling individual hardware components to form an embedded system to utilize the machine learning model. This means that the camera, neural processing unit, and central processing unit are purchased and assembled separately, requiring specific drivers and involves a lengthy setup process. Addressing these issues, we look to the usage of smartphones as a novel integrated solution for shark detection. This paper looks at using an iPhone 14 Pro as the driving force for a YOLOv5 based model, and comparing our results to previous literature in shark-based object detection. We find that our system outperforms previous methods at both higher throughput and increased accuracy.

Fitness tracking is a rapidly growing market as more people desire to take better control over their lives. And the growing availability of smartphones with sensitive sensors makes it possible for anyone to take part. This project aims to implement a Step Counter and create a model for Human Activity Recognition (HAR) to classify activities such as walking, running, cycling, ascending and descending stairs, and standing still, using sensor data from handheld devices. The Step Counter is implemented by processing acceleration data and finding and validating steps. HAR is implemented using three machine learning algorithms on processed sensor data: Random Forest (RF), Support Vector Machine (SVM), and Artificial Neural Network (ANN). The step counter achieved 99.48% accuracy. The HAR models achieved 99.7%, 99.6%, and 99.5% accuracy on RF, ANN, and SVM, respectively. / Aktivitetsspårning är en snabbt växande marknad när fler människor önskar att ta bättre kontroll över deras liv. Den växande tillgängligheten på smartphones med känsliga sensorer gör det möjligt för vem som helst att delta. Detta projekt siktar på att implementera en stegräknare samt skapa en modell för mänsklig aktivitetsigenkänning (HAR) för att klassificera aktiviteter såsom att promenera, springa, cykla, gå upp eller ner för trappor och stå stilla, med användning av sensordata från handhållna enheter. Stegräknaren implementeras genom att bearbeta accelerationsdata och hitta samt validera steg. HAR implementeras med hjälp av tre maskininlärningsalgoritmer på bearbetad sensordata: Random Forest (RF), Support Vector Machine (SVM) och Artificial Neural Network (ANN). Stegräknaren uppnådde en noggrannhet på 99.48%. HAR-modellerna uppnådde en noggrannhet på 99.7%, 99.6% samt 99.5% med RF, ANN och SVM. / Kandidatexjobb i elektroteknik 2022, KTH, Stockholm