Food–medication interactions can occur when medicines are taken with foods, drinks, herbs, or supplements that influence drug absorption, exposure, or activity. Screening these combinations is challenging because the available evidence is imbalanced, prescription text may be recognized incorrectly, unsupported inputs may produce unreliable predictions, and altered software artifacts may change the recommendation presented to the user. iMediFood-Shield addresses these concerns through an evidence-first edge-AI framework that combines structured diet–drug interaction evidence, prescription-assisted medication confirmation, coverage-aware rejection, calibrated five-class prediction, false-safe-aware confidence gating, and software-based tamper-evident verification. The DDID preparation process began with 23,950 evidence records and produced 16,644 canonical medication–food/herb pairs, including 16,165 single-effect model-eligible pairs and 479 multi-effect conflict pairs. A leakage-free 70%–15%–15% split was applied after canonicalization, and the deployed lookup was restricted to training-supported and conflict records. On the operational locked-test AI branch of 2259 supported unseen pairs, the final calibrated LinearSVC with the validation-selected MedSafe-GATE threshold of 0.65 achieved 91.72% accuracy, 80.57% balanced accuracy, and a macro F1-score of 0.8359. The gate reduced calibrated false-safe predictions from 55 to 28, corresponding to a 49.09% reduction and a final false-safe rate of 1.35% among interaction-bearing AI-branch pairs. RxOCR-Guard achieved 94.67% candidate recall and 100.00% candidate precision on a controlled synthetic prescription benchmark, while mandatory user confirmation was retained because top-1 candidate accuracy was 51.33%. The unchanged baseline and all ten adverse software-bundle conditions produced the expected verification outcomes for artifact-modification, missing-file, key-mismatch, manifest-alteration, and rollback cases. Raspberry Pi deployment reproduced all 2259 reference predictions without mismatch, completed covered AI inference in 1.737 ms on average, and verified the protected software bundle in 80.249 ms on average. These results show that iMediFood-Shield can combine evidence-grounded screening, conservative AI decision control, prescription confirmation, and software-integrity verification within a resource-constrained edge research prototype.

Falls among older adults can cause serious injury and loss of independence. cStick 2.0 is a vision-enabled, IoMT-edge based smart walking-stick prototype that combines multimodal fall-risk classification with embedded obstacle awareness. The fall-risk classifiers were evaluated using a 9670-record development dataset, on which the compact DNN achieved 95.40% accuracy, 92.35% balanced accuracy, a macro F1-score of 93.90%, and a ROC-AUC of 97.44%. The Arduino Nicla Vision obstacle module used an INT8 Edge Impulse model with centroid-based direction assignment and time-of-flight distance sensing; 144 controlled trials produced 75.00% obstacle-presence accuracy at approximately 19–20 FPS. Sensor acquisition, GPS, display output, buzzer response, and CSV record accumulation were demonstrated at a prototype level. Synchronized older-adult evaluation, device-to-application communication, secure caregiver services, multimodal accessibility feedback, and longitudinal personalization remain future validation stages.

Stress can be defined as a common psychological and physiological reaction that can appear as a result of an individual facing a demanding or challenging situation. In some situations, this stress response can be helpful in quick response to overcome the threats. But, a prolonged exposure to stress can lead to both mental and physical health issues. Physiological parameters of an individual can act as biomarkers to accurately represent the stress levels. An accurate stress detection system using these physiological signals can help in early stage detection and intervention to effective stress management. Hence, this study explores lightweight regression and classification models for stress detection on a real-time collected dataset. Regression results show that gradient boosting approaches significantly outperform other models, with Gradient Boosting, LightGBM, and XGBoost achieving test $R^{2}$ values above 0.99 and minimal prediction errors. In contrast, classification performance remains relatively low, with the best models, neural networks and boosting-based classifiers, reaching test accuracies of approximately 0.35. These findings indicate that modeling stress as a continuous variable using regression, particularly with gradient boosting methods, is more effective than discrete stress classification.

Accurately estimating calorie intake and nutrient composition from what we eat remains one of the most practical challenges in maintaining a healthy lifestyle. Manual food logging and database-based estimations are often inaccurate because ingredient proportions and preparation styles vary widely. This paper presents a lightweight, privacy-preserving framework that estimates calories and detailed nutrient values from a single image. The model uses a Mask R-CNN-based segmentation network to identify visible food components, measure their area, estimate their volume using preset height values, and map them to nutritional information obtained from reliable datasets such as USDA and Food-a-pedia. The system integrates federated learning (FL) to ensure privacy by allowing the model to improve collaboratively without sharing raw user data. The proposed architecture achieved a mean Average Precision (mAP) of 96% for detection and 92% for segmentation, confirming its precision and efficiency. The model is trained and evaluated on a curated pizza dataset consisting of 1107 images across 50 topping categories, using a standard train-validation-test split (666/219/222) to ensure reliable performance assessment. The proposed system also achieves low nutrition estimation error, with calorie and nutrient deviations remaining within approximately 3.8% to 11.1% across evaluated metrics. A lightweight mobile interface is demonstrated through a Figma-based prototype mockup to illustrate potential real-world deployment and user interaction.

Effective diet management is crucial for preventing non-communicable diseases, including heart disease, diabetes, and obesity. This paper introduces NutriVision, an enhanced system featuring an interactive chatbot that delivers real-time dietary guidance and personalized recommendations. NutriVision enables users to inquire about nutritional content, receive customized recipe suggestions, and track nutritional goals through natural language interaction. By integrating user data such as health conditions, dietary preferences, and nutritional requirements, it generates tailored responses that enhance user engagement and decision-making. With the help of computer vision and machine learning, NutriVision accurately identifies food items and estimates quantities from smartphone-captured images, providing instant nutritional analysis that includes both macronutrient and micronutrient details with a 94

Over 60% of global agricultural soils suffer from nutrient degradation, yet most farmers still lack access to real-time, affordable soil health diagnostics. To address this gap, this research presents SanaSolo 3.0, a mobile, edge-enabled, scalable, modular, and terrain-adaptive embedded system for in-situ soil fertility monitoring. Building upon previous iterations, SanaSolo 3.0 integrates a spring-suspension tracked chassis, embedded sensor suite (measuring pH, EC, moisture, temperature, and macro-nutrients), and onboard classification logic for real-time assessment. The system was validated across biologically enriched and degraded soils and compared against certified laboratory benchmarks. Results show consistent accuracy and the ability to distinguish fertility classes without invasive sampling or extensive external processing. By combining non-invasive data acquisition with biologically grounded calibration, SanaSolo 3.0 advances scalable, low-cost, and field-deployable solutions for regenerative agriculture.

It is very essential to calculate nutrients consumed for a lifestyle with health and wellness. Many existing systems and mobile applications depend on manual input by user or calculate only the dish level averages without considering multiple ingredient variations. iLog 2.1 is a lightweight and privacy preserving work which estimates the calories, protein, and carbohydrates from just a single 2D image uploaded by the user. By using a Mask R-CNN model which was trained on a custom dataset, each ingredient is detected and it’s pixel area is calculated and converted into volume using the preset height values. The nutrient information has been derived from the USDA and the Food-a-pedia datasets. The federated learning aspect ensures the privacy of the user, by sharing only the encrypted model updates. The system achieves 93% detection accuracy and 85% nutrient estimation precision, which shows efficient and secure ingredient-level analysis.

Currently, consumers identify markets to obtain necessary goods. Certain retailers offer consumers a comprehensive range of necessities, including apparel, gadgets, frozen food, etc. Fruits are considered vital commodities that should be available in supermarkets. Supermarkets enter into arrangements with large farms to acquire fruits for customer distribution; however, does the quality of the supplied fruits meet consumer satisfaction? Store owners compensate for fruits based on both quantity and quality. This challenge necessitates a system to classify the health of fruits to inform markets of the percentage of high-quality produce. This research proposes a system that classifies agricultural products by training on harvested fruits and predicting their health status. The precision of the outcomes depended on a Convolutional Neural Network (CNN) model utilizing a dataset comprising approximately 13,000 images depicting both fresh and rotten fruits.

As humans, we coexist alongside animals on our planet. Depending on this truth, adapting to them is essential for several reasons. Some people obtain their sustenance from Livestock, preserving our lives, etc. Agricultural products are crucial for humanity, and their preservation is a fundamental human responsibility. Animals are one of the threats to crop production. In this phase, it is forbidden to kill, harm, or imprison them. Simultaneously managing crops and animals is one of the most challenging jobs farmers undertake. Monitoring crop growth on the farm and protecting them from animals is an addressed issue in this paper. A system has been developed to aid farmers in safeguarding crops and preventing animal-related losses. The proposed approach is capable of deterring animals from crops without inflicting harm or damage. The proposed system utilizes You Only Look Once version 11 (YOLOv11m) along with a high-quality dataset to facilitate the classification of hazardous objects detected by the camera.

Driver distraction is a major contributor to road accidents, and the presence of children, whether crying or unusually quiet, often increases that distraction. AngelRide, an Internet of Medical Things (IoMT) system, addresses this problem by automatically monitoring a child's condition in the car seat and easing the driver's cognitive load. The device uses an advanced computer-vision pipeline: a YOLOv11 model first detects a seated child, then an eye-tracking module classifies their sleep state. High detection accuracy was achieved through the use of the YOLOv11 model and a high-quality dataset. A suite of environmental sensors continually adjusts light, temperature, humidity, and sound to maintain optimal sleeping conditions. By improving the child's comfort and minimizing driver concern, AngelRide simultaneously enhances in-vehicle safety and the child's sleep quality.

Since ancient times, humans have cared for many things-from houses and animals to farmlands. Vegetables on farms also require proper attention to remain healthy. In the past, crops were monitored manually, but rising labor costs and inflation have made that increasingly difficult. If not cared for properly, vegetables can spoil, affecting the quality of the entire harvest. This paper introduces VegeCare, a system designed to monitor vegetables, detect spoilage, and remove spoiled produce to protect the rest. VegeCare uses a trained Convolutional Neural Network (CNN) model to analyze images and assess the condition of vegetables, achieving an accuracy of 95%. The dataset used for training was sourced from Roboflow and contains over 5,000 images of both fresh and rotten vegetables. VegeCare also assists farmers and suppliers by sorting vegetables into fresh and stale categories after harvesting. This streamlines the handling process and helps ensure healthier produce reaches consumers.

In order to achieve optimal and sustainable produce, soil fertility monitoring and management are needed. A fully automated IoT-enabled edge-based system, SanaSolo 2.0, is proposed through this paper, which continuously monitors soil moisture, soil temperature, electrical conductivity (EC), pH, and the fertility levels of nitrogen (N), phosphorus (P), and potassium (K). The system is applied to two different types of soils - infertile and fertile. The vermicompost is added to the soils to observe the improvements in the nutrition content. With calibration and analyses, a significant improvement was observed when infertile soil was added with vermicompost. These findings not only demonstrate the importance and need of this system but also prove that the system is reliable, thereby supporting smart agriculture practices.

Food is crucial in our life. Despite food manufacturers' efforts to meet consumers' needs with manufactured food, it cannot attain the identical level of quality and flavor as natural food. Some fruits and vegetables are classified as food that cannot be manufactured. Also, they play an essential role worldwide. This paper focuses on how farmers can protect their farms from an object that might threaten the crops and cause damage to them. It proposes a framework that farmers may utilize to safeguard the crops against any species of birds. Considering the adverse impact of bird attacks on crop production, this system effectively addresses this problem and consistently enhances the quality and quantity of them. It utilizes computer vision technology to establish a secure environment for the crop. This system is called Quality of Crop Device (qCrop), and it works with a You Only Look Once (Yolov8m) model to detect birds with high accuracy to protect farms.

Maintaining health and fitness through a balanced diet is essential for preventing non communicable diseases such as heart disease, diabetes, and cancer. NutriVision combines smart healthcare with computer vision and machine learning to address the challenges of nutrition and dietary management. This paper introduces a novel system that can identify food items, estimate quantities, and provide comprehensive nutritional information. NutriVision employs the Faster Region based Convolutional Neural Network, a deep learning algorithm that improves object detection by generating region proposals and then classifying those regions, making it highly effective for accurate and fast food identification even in complex and disorganized meal settings. Through smartphone based image capture, NutriVision delivers instant nutritional data, including macronutrient breakdown, calorie count, and micronutrient details. One of the standout features of NutriVision is its personalized nutritional analysis and diet recommendations, which are tailored to each user's dietary preferences, nutritional needs, and health history. By providing customized advice, NutriVision helps users achieve specific health and fitness goals, such as managing dietary restrictions or controlling weight. In addition to offering precise food detection and nutritional assessment, NutriVision supports smarter dietary decisions by integrating user data with recommendations that promote a balanced, healthful diet. This system presents a practical and advanced solution for nutrition management and has the potential to significantly influence how people approach their dietary choices, promoting healthier eating habits and overall well being. This paper discusses the design, performance evaluation, and prospective applications of the NutriVision system.

To efficiently manage plant diseases, Agriculture Cyber-Physical Systems (A-CPS) have been developed to detect and localize disease infestations by integrating the Internet of Agro-Things (IoAT). By the nature of plant and pathogen interactions, the spread of a disease appears as a focus with density of infected plants and intensity of infection diminishing outwards. This gradient of infection needs variable rate and precision pesticide spraying to efficiently utilize resources and effectively handle the diseases. This article, SprayCraft presents a graph based method for disease management A-CPS to identify disease hotspots and compute near optimal path for a spraying drone to perform variable rate precision spraying. It uses graph to represent the diseased locations and their spatial relation, Message Passing is performed over the graph to compute the probability of a location to be a disease hotspot. These probabilities also serve as disease intensity measures and are used for variable rate spraying at each location. Whereas, the graph is utilized to compute tour path by considering it as Traveling Salesman Problem (TSP) for precision spraying by the drone. Proposed method has been validated on synthetic data of locations of diseased locations in a farmland.

The Internet of Things (loT) has revolutionized traditional agricultural practices, making its impact in the era of intelligent fanning. This mini-review explores the role of loTenabled electronics and smart systems and the need to deploy them to transform agriculture through real-time data collection, automation, and resource optimization. The paper also discusses the critical technologies grouped into sensors, drones, and data processing techniques for the contributions of loT-enabled electronics to precision and vertical farming, livestock monitoring, autonomous irrigation systems, soil, and fanners. The paper also highlights the benefits of loT devices by addressing global challenges such as data security and privacy, climate change, labor shortages, and sustainable energies while discussing the technical and socioeconomic challenges to widespread adoption across the agricultural community. Finally, the review further emphasizes the potential of collaborative robots, swann robots, AI and ML applications for regenerative agriculture, biometric sensors, and personalized fanning to enhance the effectiveness and scalability of loT-enabled smart agriculture.

This is an extended abstract for a Research Demo Session based on our published work [1]. PTSD has been a major problem in our society and much research has been done along the line to predict and diagnose PTSD. Our method helps to predict PTSD in its early stage with the help of physiological markers which combined with prior information about the patient like PTSD history, exposure to trauma, substance abuse disorder and other information helps to create a risk score with more accuracy. Due to the lack of a public dataset on this domain, we used different uni variate relationships of physiological markers with PTSD to create a multi-modal model using a slight modification of the naive Bayes algorithm. Implementation of a micro-controller along with the cloud IoT platform and a mobile app is created to demo the possibility of the system which helps healthcare providers and users to timely track and monitor PTSD risks with background information and priors accurately.

Worm castings (Worm Excretion) are one the richest natural fertilizers on earth, making earthworms a very important and applicable soil health indicator. According to an article published in the Polish journal of Environmental studies, the most important chemical components of worm castings are pH, total organic carbon (TOC), total nitrogen (N), plant available phosphorus (P), plant available potassium (K), and calcium water soluble (Ca). These chemical components of worm castings, paired with soil temperature, humidity and electric conductivity, are all measurable values that can indicate the overall health and fertility of soil. Furthermore, these physical-chemical properties can also be measured and analyzed to estimate worm populations in soil, making traditional manual extraction techniques obsolete. The proposed project, Sana Solo, is a device that uses machine learning to estimate worm populations based on the quantities of the physical-chemical properties listed above. Being able to estimate earthworm populations in a timely manner, without the use of extraction techniques, can be used in farms and gardens to evaluate soil fertility.

In 2020, nearly one hundred thousand deaths were drug related overdoses. In addition, forty to sixty percent of recovering addicts relapsed. A top contributor to relapsing is stress, as high levels of hormones are released within the brain. Individuals that could benefit from a rehabilitation center have many varied reasons as to why they do not seek one, proximity to a center being a primary reason. Rehabilitation facilities are a key factor in the recovery of an addict, as the probability of drug use decreases fifty to seventy percent after treatment. The hardest step in recovery is recognizing the need for help, however finding a treatment center can be a stressful experience, potentially leading to relapse. To reduce the stress of finding the closest and most qualified center, the goal of this research is to create a website that contains a database of rehabilitation centers. This could later be connected to a mobile application that would track the GPS location of the user and return the geographical location of the closest treatment center, as well as the name of the facility and contacting phone number. The software will have the ability to connect to wearable devices, allowing an opportunity to track physiological factors affected during an overdose.