AMA, Agricultural Mechanization in Asia, Africa and Latin America (AMA) (issn: 00845841) is a peer reviewed journal first published online after indexing scopus in 1982. AMA is published by Farm Machinery Industrial Research Corp and Shin-Norinsha Co. AMA publishes every subjects of general engineering and agricultural engineering.
AMA, Agricultural Mechanization in Asia, Africa and Latin America (ISSN: 00845841) is a peer-reviewed journal. The journal covers Agricultural and Biological Sciences and all sort of engineering topic. the journal's scopes are in the following fields but not limited to:Azerbaijan Medical Journal Gongcheng Kexue Yu Jishu/Advanced Engineering Science
Indonesian local coffee, widely known in the global market, is vulnerable to being counterfeited with other cheaper coffee products. Therefore, we need technology to identify the types of local Indonesian coffee. One of the non-destructive methods for identifying coffee products is computer vision. This study aimed to develop a computer vision method to classify three types of local Indonesian Arabica coffee beans. Those are Gayo Aceh, Kintamani Bali, and Toraja Tongkonan using three types of pre- trained convolutional neural network (CNN), namely GoogLeNet, SqueezeNet, and Inception-v3. Sensitivity analysis was carried out by varying the optimizer, i.e. SGDm, Adam, and RMSProp, and varying the learning rate, which included 0.00005 and 0.0001. Each type of coffee bean used 500 image data for training and validation with a ratio of 70% and 30% and 100 image data for testing. The results show that GoogLeNet, SqueezeNet, and Inception-v3 can achieve up to 100% in validation accuracy value and up to 99.67% in testing accuracy. From the results of this study, the pre-trained CNN model, SqueezeNet with an SGDm optimizer and a learning rate of 0.00005, is highly recommended to classify local Indonesian coffee beans. This is because it has the highest validation accuracy, the highest testing accuracy, the simplest CNN structure, the fastest training time, and the most stable training and validation charts.
This paper describes Try Gonna Bee, an Android-based mobile app that can be used to aid users who are enthusiastic to start keeping stingless bees (Trigona thorasica and Trigona itama species). Eager but inexperienced beekeepers may mistakenly installed bee hives in inappropriate locations due to the lack of guidance. This results in poor quality honey yield and causes the bee colony to move or die. Also, the hive’s entrance must not be situated under direct sunlight because it may melt the propolis structure and cause heat stress. Novice beekeepers may not understand how frequent harvesting, maintaining and re-queening a hive should be done. The developed app determines whether an area is suitable for a beehive installation by using the Global Positioning Satellite (GPS) information, which retrieves an image of the area through Google Maps API. It calculates the greenery percentage of the location, where higher greenery is more favorable (has better access to vegetation). As the hive’s entrance should not face direct sunlight, and the app uses sensors in the phone (gyroscope, compass, and camera) to capture images from the North, East, South, West, and the luminance detection algorithm calculates the direction with the lowest luminance. The app also has a honey tracker function to remind users of upcoming tasks. The accuracy of the app, the usability and users’ experience when using the app was evaluated in a user study with two stingless bee experts, and thirty non-trained users. Results showed that it could determine the greenery percentage better than humans, and accurately (100%) detect lowest luminance direction, compared to novice users (50%). Most of the respondents were satisfied with the usability and experience of using the application. The app can help cultivate higher interests in individuals to be involved with bee keeping, which indirectly improves food security.
This preliminary study focuses on the ammonia recovery from the organic sludge composting with a ratio of 3:7 shrimp sludge to cow dung. Clean nitrogen produced from the composting process will be used for the cultivation of high-value microalgae. One ton of compost mixture was prepared at a mix ratio of 4:15:1 sludge rice-husk compost-seed in a rotary drum reactor. Time course of CO2 from exhaust gas, compost moisture content and temperature were monitored daily. Daily percent CO2 from exhaust gas was controlled to be lower than 6% by adjusting the aeration at the range of 150 L/min to 30 L/min. The moisture content was observed to be within the range of 47 % to 62 % and the highest temperature achieved was 56.68 °C throughout 16 days of composting experiment. Exhaust gas from the reactor was channeled into a water tank containing seawater for ammonia trapping. The ammonium nitrogen, NH4+-N content in seawater was measured using indophenol blue method with a finding of 0.42 mg- NH4+-N /L at Day 16 of the composting. An additional experiment was carried out by mixing 2.6 mmol, (801 g) of calcium hydroxide, Ca (OH)2 with 50 kg of day 16 compost at 30 L/min of aeration to maximize ammonia recovery. A significant increment of ammonia concentration to 0.67 mg- NH4+-N/L was measured in seawater samples after 24 h. Our preliminary results showed relatively low NH4+-N content in seawater, thus few recommendations were suggested to optimize the composting process and maximize the ammonia recovery.
The dark brownish color effluent from palm oil mill effluent (POME) has continued to receive considerable interest and posed widespread concern due to its repercussions on water bodies. This study examines the performance of hydrophilic composite hollow fibre poly (vinylidene fluoride) (PVDF)-polyvinylpyrrolidone(PVP) membrane embedded with titanium dioxide (TiO2) nanoparticles as a robust antifouling intermediary for the decolourization of palm oil mill effluent (POME). The nanocomposite hollow fibre membrane with varied TiO2-nanoparticles loadings (0-2.0 wt.%) were swirled through phase inversion procedure to provide a negatively charged surface. The developed nanocomposite membranes were analyzed and compared according to SEM, contact angle, EDX, permeability flux, COD and color removal, and fouling resistance from (POME). Color removal, fouling resistance, and flux permeability enhanced with a rise in TiO2 loading. The flux efficiency of the membranes was assessed using pure water and POME as a feed liquid. Results show that the membrane-embedded with 1.0 wt% concentration of TiO2 exhibited higher performance showing pure water flux and POME flux, COD and color removal of 120.74 L/m2h, 74.3 L/m2h, 90.1%, and 94.7% from POME respectively. Furthermore, an excellent fouling hindrance was attained, thereby allowing the percentage of flux recovery ratio (FRR) recorded was 95.46%, 92.49%, and 87.49%, after three cycles of uninterrupted filtrations for a period of 540 min. Yet, declination in flux and fouling resistance was observed at supplemental loading of 1.5 and 2.0 wt%-TiO2 owing to the accumulation of nanofillers in the membrane matrix at the higher concentrations.
This study provides the use of a near-infrared (NIR) sensor with an appropriate gimbal mounted on a lightweight Unmanned Aerial Vehicle (UAV) to operate remote sensing acquisitions of oil palm trees between the age of 9 to 15 years, at low altitudes. UAVs and NIR sensor calibrations and corrections were used in this study to develop a correction approach for obtaining accurate remote sensing images that may be used to monitor oil palms. The major aspects of calibration required for quality control of the flight data are UAV attitude flight information such as heading, pitch, and yaw. Compared to the previous system, which is based on a gyroscopic instrument, this flight information delivers superior precision and reliability. Geometric and radiometric sensor corrections and calibrations were carried out. After correcting undesired and sensor fault features, reliability image data analysis was created. As a new approach, the Improved Normalized Difference Vegetation Index (INDVI) algorithm uses the red and green channels as reflectance and the blue channel as absorption (400-1100 nm region), where healthy plants normally reflect green light in NIR and visible light, as opposed to the traditional NDVI, which only uses the near-infrared and red channels where the processing image is converted using ImageJ software. The results only discuss whether an area is rich in biomass or has less vegetation, which influences the results because INDVI and NDVI are two different indices.