This report details the development of a device that utilises image detection and machine learning to predict slug infestation in greenhouses. Increasing pesticide regulations have led to a rise in slug infestations, causing significant crop damage. Although alternatives to pesticides exist, they are expensive and labour-intensive to implement across large greenhouses. The goal of this project is to reduce labour by automating slug monitoring through image detection and machine learning. The focus is narrowed to detecting the Spanish Earthslug (Lehmannia Valentiana) in Cut Cymbidium, a type of orchid, as this market is particularly affected, and one slug species makes image detection more manageable. Context research and experiments were done to understand the problem. Slugs are nocturnal creatures that emerge from the soil periodically for food. Their activity is mainly influenced by temperature and humidity, preferring moist environments at around 17°C. Since greenhouse conditions vary by plant growth stages and change throughout the year, it is worthwhile to monitor activity on a longer time scale. Image detection and computer vision are already used in pest control and other applications within greenhouses. For efficient image detection, the looked after objects must be clearly identifiable and the images should preferably be of consistent quality. To achieve this in a greenhouse a good suggestion would be a controlled environment with minimal to no leaf occlusion. In our case, another important factor is having low data images with a high contrast between the slug and their background. This can be achieved by adjusting lighting conditions, converting images to black and white, and tracking changes between images to detect slug movement. After these results, the direction was chosen to predict slug infestation based on environmental data using counted slugs through image detection as a reference. Based on these insights, a device concept was developed using a Seeeduino Xiao ESP32S3 Sense microcontroller and Grove sensors. The system lures slugs with wheat bran, a proven bait, and uses a camera to capture images. Photos are stored on an SD card along with environmental measurements, such as temperature and humidity. Field tests and prototyping demonstrated that a 12 cm-high container with four 1 cm-diameter entryways was effective for attracting slugs and capturing quality images. To enhance contrast and prevent mould growth, a white mesh was placed between the bait and the slugs. For easy processing and low data images, the images are processed into binary images, where pixels are either black or white, and Hough Circle transform is used to crop the images around detected circles. For slug population prediction, a regression algorithm was selected, with environmental variables serving as the input and slug counts as the output labels. Spatial interpolation can be used to estimate environmental values for the entire greenhouse based on measurements of the device therefore getting the data to make predictions for slug infestation. The final version of the device, called Cephal, was tested in the field and used to build an image detection model. During field tests, Cephal successfully captured environmental data and collected over 485 images, although no slugs were detected, indicating the need for further research. Additionally, improvements such as a longer cable for the moisture sensor and a more accurate DHT sensor are required. With Cephal, a dataset of 482 images was used for machine learning. Reducing the input size to 64x64 binary images, a CNN model was made with around 90% accuracy. In conclusion, Cephal has the potential to enhance slug monitoring in greenhouses. With improvements in sensor performance and further optimisation of the machine learning model, Cephal could accurately predict slug infestation, helping greenhouse owners implement timely pest control strategies. Future work should focus on optimising the electronics and training and optimising the slug prediction model based on environmental values.

This graduation project, conducted in collaboration with Agis Innovations, aimed to develop a cow localisation solution that could be integrated with existing ear sensors for cows. This addition was crucial as current cow localisation solutions were inaccurate, expensive, or not user-friendly. The inability to promptly locate cows can cause unnecessary pain for the animals and reduce farm efficiency. Moreover, timely cow localisation can reduce the need for fertility hormones and antibiotics.

The project was divided into two main aspects: technology and user interaction, allowing for separate evaluation and design optimisation. A comprehensive literature review on cow localisation technologies and the existing market revealed the need for a solution that is highly accurate, cost-effective, supports long battery life, and offers a long detection distance. No existing solutions met all these criteria. The technology analysis identified an unused, novel, and promising new distance measuring technology. User research highlighted the shift from small to large farms, requiring a larger, often low-literacy workforce. Literature and empirical data from farmer interviews and worker questionnaires pointed to low work engagement, particularly in cow localisation tasks. Consequently, two user evaluation criteria were established: intuitive user touchpoints and a user interface (UI) that enhances work engagement.

From the user perspective, analysing the complete workflow of cow localisation revealed three phases, each having different detection distances and accuracies. Integrating these phases provided a solution with both a long detection distance and high accuracy. Insights into farm workers’ engagement revealed unmet fundamental needs. By incorporating personal profiles, hands-on tutorials, and gamification features like leaderboards and rewards, the UI was designed to enhance workers’ sense of competence, recognition, and purpose. This improved engagement is anticipated to enhance cow care and increase productivity. The UI’s intuitiveness was based on the three localisation phases, with each phase having tailored visualisations to guide users towards the cows effectively.

In the technology aspect, the new technology was first evaluated through field testing and Python simulations to prototype localisation ideas. A physical prototype demonstrated the feasibility of locating a tag device, achieving an accuracy within 2 metres. This met the accuracy requirement, allowing users to identify cows easily. The limited use of routers kept costs low, while smart location update frequencies minimised power consumption, supporting long battery life.

Integrating the user and technology aspects addressed the research question. The localisation solution is feasible, requiring minimal hardware changes for Agis Innovations. It is viable, leveraging the latest technology with a user-centred development approach, and desirable, providing an essential tool for efficient herd management. This solution not only boosts productivity but also enhances cow wellbeing, aligning with the project’s ultimate goal of improving animal welfare.

This project aims to create a proof of concept to evaluate the potential value of implementing a heat recovery system for Quooker in the kitchen context. In response to the global goal of achieving net-zero CO2 emissions by 2050, reducing energy loss has become a shared priority. Heat recovery systems can significantly decrease energy consumption, leading to economic and sustainable benefits. In collaboration with Quooker B.V., this project assesses the feasibility and impact of integrating such a system in kitchen environments. The evaluation follows a triple diamond approach, consisting of four stages: discover, define, develop, and deliver. During the discover phase, extensive research is conducted on kitchen contexts, user behavior, existing heat recovery systems, and the characteristics of current kitchen setups. The define phase involves exploring various design directions, identifying potential hazards, and selecting a single promising direction for further development. In the development phase, two design principles are prototyped and tested to determine which one offers the optimal heat recovery rate relative to its initial cost and carbon footprint. The deliver phase focuses on final testing, where the developed concept is evaluated for efficiency during typical kitchen activities, such as draining water after cooking and using the dishwasher. The results are analyzed over a 10-year period of daily use. The project concludes with an assessment of the final concept’s viability, feasibility, and desirability. Recommendations for further research and potential hazards associated with the implementation of
the heat recovery system for Quooker are provided.

This report concerned the development of a soft robotic module for a seat pan, its optimization and application. The soft robotic module has an LED and photosensors for determining the distance of indentation and the passive force created by compressing the foam spring inside. An air pressure sensor is used to determine the active force created when an air pump inflates the bellow with air. The system is trained by machine learning to calculate predictions of these distance and forces in real-time. In several iterations of the module the reproducibility and accuracy were developed in such a way that it could be built into a seat. Two modules are built into a seat pan and interestingly participants on the seat were able to experience significant comfort differences, showing that the principle works. Further development is needed to make a seat pan with more modules, combined with a central computing system that monitors, records and regulates the modules. Exploration of simplification by using a linear regression instead of a neural network to calculate the active force is recommended, as well exploration of improved functionality by dividing the neural network that calculates the distance of indentation.

The goal of this Master thesis was preventing the onset of Repetitive Strain Injury (RSI), Computer Vision Syndrome (CVS) and technostress among computer workers through encouraging them to take more breaks at work. The project was executed in a collaboration with TU Delft and R-Go Tools and took approximately six months. RSI, CVS and technostress affect the personal as well as the professional happiness of computer workers. Among different preventions, aimed at decreasing the risk of developing these syndromes, implementing frequent breaks at work, especially macro breaks, seems to be very promising. Unfortunately, the existing anti-RSI software, aimed at facilitating breaks at work, is considered to be annoying and disturbing. Therefore, the majority of computer workers still take an insufficient number of breaks, putting themselves at a high risk of developing the aforementioned chronic diseases. The analysis of the problem included literature review, company analysis, user research and context analysis. Afterwards, different ideas addressing the problem were generated based on the formulated list of requirements. Consequently, the best ideas were developed into concepts. In the end, Break Bottle was chosen as the most suitable solution for the assigned problem. This concept is a smart bottle which encourages users to drink water, to refill the bottle and, eventually, to go to the toilet. The signals come from changes in the appearance of the object. Consequently, the users are expected to take three different macro breaks. The lights were chosen for communicating with the user. Different light sequences are supposed to trigger different responses from the user. The lights are placed inside the shell behind transparent indents. The form of the bottle enables firmer grasping and holding of the object. The grey paint and the green lights together make the bottle look either lively or lifeless depending on whether the lights are switched on or off. Moreover, the product fits well in the office environment, presenting itself both sleek and elegant. The existing anti-RSI software of R-Go Tools needs to be adjusted for controlling the bottle. The type of signal coming from the bottle depends on whether the user drinks the sufficient amount of water and takes enough breaks. A crucial part of the software is a seduction algorithm whose aim is to indirectly make users take more breaks at work. The main objective of the project has been met since the designed product facilitates taking bre¬aks at work by computer workers. The annoyance with the product is expected to be lower than with the existing anti-RSI software since the users already accept the proposed breaks. Moreover, the product facilitates taking macro breaks which increases the chance of preventing the onset of RSI, CVS and technostress. Since the bottle looks very appealing, it is foreseen that the users will be willing to interact with it. Therefore, it is expected that computer workers behaviour will change in response to the appropriate signals coming from the product.

This master thesis analyses and proposes a design for a direct air capture (DAC) unit as part of a small scale, solar powered methanol synthesis plant. This project analyses the existing process and design by ZEF BV and explores solutions for mass manufacturing a direct air capture unit. As a result a concept design solution is found and analyzed on cost and environmental impact. To achieve this, several product design methods have been applied, including idea generation and prototyping, all in an iterative process. Sub-solutions have been designed an tested and are brought together in a system architecture to form a product. A direct air capture unit provides an access point to capture CO2 and water from air and turn air into methanol with the power of sunlight. A viable product design can be achieved by integrating product functions in single components rather than making components fulfill a single function. In this way,
resources can be used effectively. This thesis shows that it is possible to make such a product, however, many aspects leave room for improvement or may need to change depending on other the needs of other systems of a small scale methanol synthesis plant. It provides a base for a next generation of DAC design.

One of the major challenges that low income countries have been facing for decades is to attain food security. The project covered in this report investigates how solar adsorption refrigeration technology can be used to achieve a higher level of food security in West-Africa. More specifically, the study aims to conceptualize suitable product proposals that integrate the technology in order to improve food security, be profitable and provide a more environmentally sustainable alternative to current food conservation practices. To this end, the principle of food security is explored. Technological and contextual research
is subsequently executed. Findings are incorporated in the design phase during which 3 concepts are proposed and evaluated. The most promising concept, the Solar Stall, is detailed in the embodiment phase.

“Food security, exists when all people, at all times, have physical, social and economic access to sufficient, safe and nutritious food that meets their dietary needs and food preferences for
an active and healthy life” (US Global Change Research Program et al., 2015). Solar energy based refrigeration means can improve food security in two manners. Firstly, solar powered refrigeration presents the opportunity to adequately preserve produce, extending shelf life and therefore reducing produce loss due to spoilage. A reduction of produce loss in turn provides a higher level of food security. Secondly, solar powered refrigeration can enable an
increase in domestic food production, which has a major impact on food security. These two components are leading parameters throughout research and conceptualization.

Solar adsorption refrigeration utilizes the processes of adsorption and evaporative cooling to provide refrigeration. These processes are driven by temperature fluctuations within the system, which are instigated through solar heat capture. The main advantage of solar adsorption refrigeration is that it has the potential to be very cost effective, making it particularly suitable for low income coutries. Its main disadvantage is a relatively restricted efficiency caused by poor utilization of heat to trigger
adsorption. The underlying reasons are
addressed within the technology section of this report.

The state of food security differentiates among all nations in West-Africa. The contextual research that was performed during this project therefore looks at both Ivory Coast, which is considered to exhibit a high degree of food security, and Liberia, which is considered to exhibit a low degree of food security. An understanding of these extremes allows to establish a spectrum of data that is representative for the entire West-African region. The value chains of perishable goods of both countries were analysed. Based on early findings, the research
is narrowed to specifically consider the retail link of the value chain, where produce is sold in open-air markets. The research concludes that the stimulation of domestic food production in countries with a state of food security similar to Liberia, should be the main objective of
the product proposal. Furthermore, the research points towards fish and meat retailers in the open-air markets, which supply the general public with their daily groceries, as the most beneficial target group.

Three suitable concepts have been developed and evaluated based on their impact in terms of food security, profitability and environmental sustainability. The Solar Stall substitutes the current market stall and incorporates a cooled compartment for produce conservation. This concept was considered to comply best with the specified requirements and was therefore
detailed during the final stage of the project.