http://localhost:8080/xmlui/handle/123456789/5 Sun, 13 Sep 2026 19:13:51 GMT 2026-09-13T19:13:51Z http://localhost:8080/xmlui/handle/123456789/2422 DEVELOPMENT OF A BIRD DETECTION AND DRONE BASED DETTERENT SYSTEM APARNA K; ALEESHA K S; HIBA K V; Dr. NISHANTH M STANLY, (Guide) Bird damage to paddy crops represents one of the most persistent and economically significant biotic stressors affecting agricultural productivity in Kerala, particularly during the heading and grain filling stages. The Baya Weaver bird (Ploceus philippinus), locally known as Attakili, is among the most destructive avian pests in the paddy growing regions. Conventional bird scaring methods including scarecrows, reflective tapes, gas cannons, and manual patrolling suffer from rapid habituation declining rural agricultural labour availability. This study presents the prototype development and evaluation of bird deterrent system, an integrated nano drone based bird deterrent system designed for small and marginal paddy farmers in Kerala. The system employs a two station architecture comprising a ground station and a drone mounted deterrent station. The ground station utilises an ESP32-CAM module equipped with an OV3660 image sensor for continuous field surveillance. Bird detection is performed using the Gaussian Mixture Model (GMM) algorithm,Upon detection, the system simultaneously dispatches a real time alert to the farmer's mobile phone via the Telegram notification platform and, after a ten second delay, wirelessly transmits a 433 MHz RF signal to the drone mounted receiver unit.The drone unit, comprising an Arduino Nano microcontroller and a piezoelectric buzzer, activatesan acoustic deterrent upon receiving the signal. Comparative evaluation of GMM and K-Nearest Neighbours (KNN) background subtraction algorithms was performed across eleven probability threshold values (0.30 to 0.99) using a ground truth dataset of 30 bird occurrences. The GMM achieved a consistent True Positive Rate of 0.900 across all thresholds, with the False Positive Rate declining monotonically from 0.739 at threshold0.30 to 0.423 at threshold 0.99. The KNN model produced a fixed True Positive Rate of 0.767 and an invariant False Positive Rate of 0.906 across all thresholds, indicating threshold insensitivity and substantially higher false alarm generation. GMM operated at threshold 0.99 was selected as the optimum detection configuration. Field deterrence trials over six observation sessions demonstrated that bird deterrent system achieved a mean bird return time of 17.00 ± 2.00 minutes,compared to 2.99 ± 0.52 minutes for conventional manual scaring a 5.7-fold improvement in deterrence duration. The bird deterrent system demonstrates that low cost embedded IoT components, combined with motion based computer vision and aerial acoustic deterrence, can deliver effective, scalable, and farmer accessible bird management for paddy cultivation in Kerala. Thu, 10 Sep 2026 00:00:00 GMT http://localhost:8080/xmlui/handle/123456789/2422 2026-09-10T00:00:00Z http://localhost:8080/xmlui/handle/123456789/2417 KINEMATIC DESIGN OF ROBOTIC BLACK PEPPER HARVESTER DONA DOMINIC; AKHINA V S; ASWIN P; Sindhu Bhaskar, (Guide) Thu, 01 Jan 2026 00:00:00 GMT http://localhost:8080/xmlui/handle/123456789/2417 2026-01-01T00:00:00Z http://localhost:8080/xmlui/handle/123456789/2415 BANANA DE-RATOONING MACHINE ASNA SHERIN; NISHANA M P; SAFANA C; Rajesh A N, (Guide) Thu, 01 Jan 2026 00:00:00 GMT http://localhost:8080/xmlui/handle/123456789/2415 2026-01-01T00:00:00Z http://localhost:8080/xmlui/handle/123456789/2412 DEVELOPMENT AND EVALUATION OF SELF- PROPELLED DRUM SEEDER MISRIYA K; AMNA K; FEMINA BANU K C; Edwin Benjamin, (Guide) Rice cultivation is increasingly affected by labour scarcity, rising production costs, water stress, and the limitations of conventional transplanted rice systems. Direct seeded rice (DSR) has emerged as a promising alternative due to reduced labour requirement, lower water consumption, and timely crop establishment. Among DSR methods, drum seeding of pre-germinated paddy is widely practiced, but the conventional manually operated drum seeder causes operator fatigue and has limited field capacity. To overcome these limitations, the present study entitled “Development of Self-Propelled Drum Seeder” was conducted at the Department of Farm Machinery and Power Engineering, KCAEFT, Tavanur. The study aimed to determine the physical and engineering properties of paddy seeds, optimize seeding parameters under laboratory conditions, develop a self-propelled drum seeder, and evaluate its field performance. Two rice varieties, Jyothi and Uma, were selected for the study. A laboratory test rig with a sticky belt mechanism driven by a variable frequency drive was fabricated to simulate different forward speeds. A factorial experiment was conducted to study the effects of variety, hole diameter, hole spacing, and forward speed on seeding performance parameters. The results showed that seeding performance was significantly influenced by the interaction of these factors. Among the combinations tested, the Uma variety with 9 mm hole diameter, 80 mm hole spacing, and 1.5 km/h forward speed was identified as the optimum configuration based on overall seeding performance. Based on these optimized parameters, a self- propelled drum seeder was developed by integrating a walk-behind power unit fitted with a 1.5 hp two-stroke petrol engine with a manually operated eight-row drum seeder. Field evaluation of the developed machine under puddled conditions showed an effective field capacity of 0.17 ha/h, field efficiency of 70.83%, fuel consumption of 0.28 L/h, and actual seed rate of 54.3 kg/ha. The study concluded that the developed self-propelled drum seeder effectively reduced operator drudgery while maintaining efficient seed metering, making it a practical and low-cost mechanized solution for wet direct seeding of paddy in small and fragmented fields Thu, 01 Jan 2026 00:00:00 GMT http://localhost:8080/xmlui/handle/123456789/2412 2026-01-01T00:00:00Z