Browsing by Title

Insect pests substantially impact global agriculture, and pest control is essential for global food production. However, some pest control measures, such as intensive insecticide use, can have adverse ecological and economic effects. Consequently, there is a growing need for advanced pest management tools that can be integrated into intelligent farming strategies and precision agriculture. This study explores the potential of a machine learning tool to automatically identify and quantify fruit fly pests from images in the context of Ghanaian mango orchards in West Africa. Fruit flies provide a special challenge for computer vision-based deep learning due to their small size and taxonomic diversity. Insects were captured using sticky traps together with attractant pheromones. The traps were then photographed in the field using regular smartphone cameras. The image data contained 1434 examples of the targeted pests, and it was used to train a convolutional neural network model (CNN) for counting and classifying the fruit flies into two different genera: Bactrocera and Ceratits. High-resolution images were used to train the YOLOv7 object detection algorithm. The training involved manual hyper-parameter optimization emphasizing pre-selected hyper parameters. The focus was on employing appropriate evaluation metrics during model training. The final model had a mean average precision (mAP) of 0.746 and was able to identify 82% of the Ceratitis and 70% of the Bactrocera examples in the validation data. Results promote the advantages of a computer vision-based solution for automated multi-class insect identification and counting. Low-effort data collection using smartphones is sufficient to train a modern CNN model efficiently, even with a limited number of field images. Further research is needed to effectively integrate this technology into decision-making systems for pre cision agriculture in tropical Africa. Nevertheless, this work serves as a proof of concept, show casing the serious potential of computer vision-based models in automated or semi-automated pest monitoring. Such models can enable new strategies for monitoring pest populations and targeting pest control methods. The same technology has potential not only in agriculture but in insect monitoring in general.

Atomic force microscopy (AFM) is a widely utilized characterization method capable of capturing atomic level detail in individual organic molecules. However, an AFM image contains relatively little information about the deeper atoms in a molecule and thus interpretation of AFM images of non-planar molecules offers significant challenges for human experts. An end-to-end solution starting from an AFM imaging system ending in an automated image interpreter would be a valuable asset for all research utilizing AFM. Machine learning has become a ubiquitous tool in all areas of science. Artificial neural networks (ANNs), a specific machine learning tool, have also arisen as a popular method many fields including medical imaging, self-driving cars and facial recognition systems. In recent years, progress towards interpreting AFM images from more complicated samples has been made utilizing ANNs. In this thesis, we aim to predict sample structures from AFM images by modeling the molecule as a graph and using a generative model to build the molecular structure atom-by-atom and bond-by-bond. The generative model uses two types of ANNs, a convolutional attention mechanism to process the AFM images and a graph neural network to process the generated molecule. The model is trained and tested using simulated AFM images. The results of the thesis show that the model has the capability to learn even slight details from complicated AFM images, especially when the model only adds a single atom to the molecule. However, there are challenges to overcome in the generative model for it to become a part of a fully capable end-to-end AFM process.

Around the world, cities are using branding as a discursive and strategic practice to adjust to intensified, ongoing competition of tourists, investments, events and skilled labour. Simultaneously, in the era of the societal turning point, sustainability issues have become a global topic, and cities have begun to brand themselves as ‘pioneer’ in sustainability. Gradually, place branding’s potential as a strategic instrument of urban development and change has been understood, and therefore, it is increasingly applied in urban governance. This thesis focuses on this change in place branding and explores the relationship between place branding and sustainable development in the context of Helsinki’s branding. More specifically, I study how place branding can be harnessed as a transformative and strategic tool to further sustainable urban development. The theoretical foundation is built on place branding literature that takes into consideration the diverse and transformative role of place branding. I reinforce the place branding theory with the concept of imaginary, which are visions of the future utilised to steer decision-making and further policies. The imaginaries can act as technologies of governance, through which cities delegate responsibility for the citizens to guide them towards a specific aim, for instance, ‘Sustainable Helsinki’. My research data consists of strategies and a website produced by the City of Helsinki. The material addresses sustainable development and the City’s branding cuts through all content. I analyse the content through frame analysis to find how Helsinki frames itself in terms of sustainable development and if any imaginaries attempt to steer the citizens to take responsibility for their sustainability actions. My research findings confirm the increasingly common perception in place brand research according to which place branding can be used as a comprehensive strategic tool in urban development. In Helsinki, place branding has moved over from mere city marketing towards a governance strategy whose objective is to both manage perceptions about places and shape the place according to the city strategies or policies. Also, what stood out was the emphasis on economic sustainability, which was visible even in sections that addressed the other two dimensions – environmental or social. This finding highlights how Helsinki’s branding is heavily influenced by the common narratives of economic success and international competition. Central findings in my research were that Helsinki uses competitive and cooperative ways of portraying itself in sustainable development and succeeding in global competition. In both of these frames, Helsinki uses imaginaries of ‘Sustainable Helsinki’, but in different ways. In the competitive tone of voice, the delegation of responsibility is more implying and indirect since the focus is on the objective, not the process. In cooperative framing, the imaginaries are more straightforwardly asserting responsibility to people and businesses. My research shows that there are several ways to guide people through place branding, but in Helsinki’s case, the city is appealing to the freedom and independence of its locals.

This master’s thesis is a case study about the perspectives that the mobile youth in the Nairobi region have in their roles in the changing world. The study begins with an assumption that the hegemonic narrative of the modern nation-state is being challenged in the globalizing world. It focuses on the worldviews and identities of the youth who have previously been awarded with a scholarship in the United States via Kennedy-Lugar Youth Exchange and Study (YES) program and currently influence in the Nairobi region. Place and citizenship are used as central theoretical concepts in the study. A multimethod design of qualitative approaches was used to conduct the empirical part of the research. The primary data was collected from the Nairobi region YES alumni by semi-structured interviews, observation and a focus group session. In addition, an expert interview was conducted and public reports regarding the subject of the research were used as secondary data. The data was analyzed with Atlas.ti software, combining coding and qualitative content analysis. The main findings of the study state that the YES alumni are globally oriented mobile and flexible citizens who identify with multiple groups and places. They sense belonging to Kenya but also identify as global citizens. Values that emerge during the research are especially related to learning and experiencing, benevolence and being successful. The youth aim to develop skills that respond to the challenges presented by globalization. In addition, the participants describe the exchange experience in the United States as an important factor that has changed the course of their lives in one way or another. As a conclusion it can be summed that the Nairobi region YES alumni are in a position where adjusting to to the changing world is possible especially due to tertiary education and possessing skills that maybe be applied in transnational expert careers. In the end of the research it is suggested that the groups who do not adjust to the changes as flexibly or whose worldviews and identities do not match in the narratives of globalization and mobility should be acknowledged next.

Pienet molekyylit (hapteenit) eivät normaalisti ole antigeenisiä. Niille voidaan tuottaa vasta-aineita liittämällä ne suureen kantaja-molekyyliin, jolloin elimistön immuunipuolustus tunnistaa ne. Vasta-aineita voidaan käyttää molekyylien tunnistamiseen solutasolla (immunosytokemia). Ligniini on luonnon toiseksi yleisin puupolymeeri, joka tukee ja tilkitsee kasvien soluseiniä. Se on lähinnä kolmesta prekursorista muodostuva fenolinen polymeeri, jonka rakenne on epäsäännöllinen ja vaikeasti tutkittavissa oleva. Ligniinin rakennetta voidaan tutkia immunosytokemiallisesti vasta-aineiden avulla. Kirjallisuuskatsauksessa käydään läpi puun ja ligniinin rakennetta, yleisiä hapteenikonjugaattien synteesimenetelmiä, ligniinirakenteita joille on tuotettu vasta-aineita sekä vasta-aineiden käyttöä erilaisten ligniinirakenteiden paikantamisessa eri kasvilajien soluseinissä. Ligniinirakenteita on havaittu eri kasvilajeilla etenkin soluseinän sekundaariseinämässä. Tutkielman kokeellinen osa keskittyy neljän dimeerisen ligniinin malliyhdisteen (hapteenin) proteiinikonjugaattien synteesiin. Valitut hapteenit edustavat yleisimpiä ligniinin rakenneyksiköitä (β-O-4', β-β', β-5'). Aryyliglyseroli-β-aryylieetterit syntetisoitiin nk. Nakatsubon menetelmää käyttäen. Resinolirakenne (β-β') syntetisoitiin entsymaattisesti. Myös β-5'-rakenne yritettiin syntetisoida entsymaattisesti. Hapteenit liitettiin proteiiniin karboksyylihappolinkkeristä aktiivisen esterin kautta. Valmiit hapteeni-proteiini –konjugaatit analysoitiin MALDI-TOF –massaspektrometrialla. Kolme hapteeni-proteiini –konjugaattia syntetisoitiin onnistuneesti. Fenyylikumaraanirakenteen (β-5') proteiinikonjugaatin synteesi ei onnistunut. Sitä voidaan kokeilla tulevaisuudessa mm. Sonogashira-reaktion avulla.

Numerical weather prediction models are the backbone of modern weather forecasting. They discretise and approximate the continuous multi-scale atmosphere into computable chunks. Thus, small-scale and complex processes must be parametrised rather than explicitly calculated. This introduces parameters estimated by empirical methods best fit the observed nature. However, the changes to the parameters are changing the properties of the model itself. This work quantifies the impact parameter optimisation has on ensemble forecasts. OpenEPS allows running automated ensemble forecasts in a scientific setting. Here, it uses the OpenIFS model at T255L91 resolution with a 20 min timestep to create 10-day forecasts, which are initialised every week in the period from 1.12.2016 to 30.11.2017. Four different experiments are devised to study the impact on the forecast. The experiments only differ in the parameter values supplied to OpenIFS, all other boundary conditions are held constant. The parameters for the experiments are obtained using the EPPES optimisation tool with different goals. The first experiment minimises the cost function by supplying knowledge regarding the ensemble initial perturbation. The second experiment takes a set of parameters with a worse cost function value. Experiments three and four replicate experiments one and two with the difference that the ensemble initial perturbations are not provided to EPPES. The quality of an ensemble forecast is quantified with a series of metrics. Root mean squared error, spread, and continuous ranked probability score are used with ERA5 reanalysis data as the reference, while the filter likelihood score is providing a direct comparison with observations. The results are summarised in comprehensive scorecards. This work shows that optimising parameters decreases the root mean square error and continuous ranked probability score of the ensemble forecast. However, if the initial perturbations are included in the optimisation the spread of the ensemble is strongly limited. It also could be shown that this effect is reversed if the parameters are tuned with a worse cost function. Nonetheless, when excluding the initial perturbations from the optimisation process, then a better model can be achieved without sacrificing the ensemble spread.

This thesis is about learning the globally optimal Bayesian network structure from fully observed dataset, by using score-based method. This structure learning problem is NP- hard, and has attracted the attention of many researchers. We first introduce the necessary background of the problem, then review various score-based methods and algorithms proposed in solving the problem. Parallelization has come under the spotlight during recent years, as it can utilize shared memory and computing power of multi-core supercomputers or computer clusters. We implemented a parallel algorithm Para-OS, which is based on dynamic programming. Experiments were performed in order to evaluate the algorithm. We also propose an improved version of Para-OS, which separates the scoring phase totally from the learning phase, performs score pruning by using Sparse Parent Graph, in addition largely reduces the communication between processors. Empirical results shows the new version saves memory comparing to Para-OS, and provides good runtime with multi-treading.

The cloud computing paradigm has risen, during the last 20 years, to the task of bringing powerful computational services to the masses. Centralizing the computer hardware to a few large data centers has brought large monetary savings, but at the cost of a greater geographical distance between the server and the client. As a new generation of thin clients have emerged, e.g. smartphones and IoT-devices, the larger latencies induced by these greater distances, can limit the applications that could benefit from using the vast resources available in cloud computing. Not long after the explosive growth of cloud computing, a new paradigm, edge computing has risen. Edge computing aims at bringing the resources generally found in cloud computing closer to the edge where many of the end-users, clients and data producers reside. In this thesis, I will present the edge computing concept as well as the technologies enabling it. Furthermore I will show a few edge computing concepts and architectures, including multi- access edge computing (MEC), Fog computing and intelligent containers (ICON). Finally, I will also present a new edge-orchestrator, the ICON Python Orchestrator (IPO), that enables intelligent containers to migrate closer to the users. The ICON Python orchestrator tests the feasibility of the ICON concept and provides per- formance measurements that can be compared to other contemporary edge computing im- plementations. In this thesis, I will present the IPO architecture design including challenges encountered during the implementation phase and solutions to specific problems. I will also show the testing and validation setup. By using the artificial testing and validation network, client migration speeds were measured using three different cases - redirection, cache hot ICON migration and cache cold ICON migration. While there is room for improvements, the migration speeds measured are on par with other edge computing implementations.

Background: Continuous practices are common in today’s software development and the terms DevOps, continuous integration, continuous delivery and continuous deployment are fre- quently used. While each one of the practices helps in making agile development more agile, using them requires a lot of effort from the development team as they are not only about au- tomating tasks but also about how development should be done. Out of the three continuous practices mentioned above, continuous delivery and deployment focus on the deployability of the application. Implementing continuous delivery or deployment is a difficult task, especially for legacy software that can set limitations on how these practices can be taken into use. Aims: The aim of this study is to design and implement a continuous delivery process in a case project that does not have any type of automation regarding deployments. Method: Challenges of the current manual deployment process were identified and based on the identified challenges, a model continuous delivery process was designed. The identified challenges were also compared to the academic literature on the topic and solutions were taken into consideration when the model was designed. Based on the design, a prototype was created that automates the deploy- ment. The model and the prototype were then evaluated to see how it addresses the previously identified challenges. Results: The model provides a more robust deployment process, and the prototype automates most of the bigger tasks in deployment and provides valuable information about the deployments. However, due to the limitations of the architecture, only some of the tasks could be automated. Conclusions: Taking continuous delivery or deployment into use in legacy software is a difficult task, as the existing software sets a lot of limitations on what can be realistically done. However, the results of this study prove that continuous delivery is achievable to some degree even without larger changes to the software itself.

Low-level wind shear is a significant aviation hazard. A sudden reduction in the headwind along an aircraft flight path can induce a loss of lift, from which an aircraft may not be able to recover when it is close to the ground. Airports therefore use low-level wind shear alert systems to monitor wind velocities within the airport terminal area and alert of any detected hazardous wind shear. There exist three ground-based sensor systems capable of independently observing low-level wind shear: a Doppler weather radar-based, a Doppler wind lidar-based, and an anemometer-based system. However, as no single sensor system is capable of all-weather wind shear observations, multiple alert systems are used simultaneously, and observations from each system are integrated to produce one set of integrated wind shear alerts. Algorithms for integrating Doppler weather radar and anemometer wind shear observations were originally developed in the early 1990s. However, the addition of the Doppler wind lidar-based alert system in more recent years warrants updates to the existing radar/anemometer integration algorithms. This thesis presents four different replacement candidates for the original radar/anemometer integration algorithms. A grid-based integration approach, where observations from different sensor systems are mapped onto a common grid and integrated, is found to best accommodate central integration considerations, and is recommended as the replacement to the original radar/anemometer algorithms in operational use. The grid-based approach is discussed in further detail, and a first possible implementation of the algorithm is presented. In addition, ways of validating the algorithm and adopting it for operational use are outlined.