Workshop Project Report 

Project Information

Rationale

Livestock industry and aquaculture are very important agricultural sectors for most of the Asian and Pacific countries.  Animal and fish products are also the important protein source for people in the region. In recent years, these two sectors are facing the challenges of increasing demand for safe and nutritious food, labor shortage and aging farmers and fish folks, climate change, emerging animal and fish diseases, etc. Government and consumer awareness on sustainable production poses another challenge to the livestock and fisheries industries from a new perspective. These issues require radical and rapid changes to the existing production methods.

In order to cope with these challenges, it has become one of the key priorities of many countries in the region to promote the research and development (R&D) and adoption of ICT technologies such as sensors, IoTs, Big Data, Artificial Intelligence (AI), agricultural robotics in improving the production efficiency, strengthening biosecurity in farms, and reducing farmers and fish folks’ workload as well as sustainability. 

Livestock industry and aquaculture are very important agricultural sectors for most of the Asian and Pacific countries.  Animal and fish products are also the important protein source for people in the region. In recent years, these two sectors are facing the challenges of increasing demand for safe and nutritious food, labor shortage and aging farmers and fish folks, climate change, emerging animal and fish diseases, etc. Government and consumer awareness on sustainable production pose another challenge to the livestock and fisheries industries from a new perspective. These issues require radical and rapid changes to the existing production methods.

In order to cope with these challenges, it has become one of the key priorities of many countries in the region to promote the research and development (R&D) and adoption of ICT technologies such as sensors, IoT, Big Data, Artificial Intelligence (AI), agricultural robotics in improving the production efficiency, strengthening biosecurity in farms, and reducing farmers and fish folks’ workload as well as sustainability. 

Currently, large scale livestock and aquaculture industries in the Asian and Pacific region have been actively introducing the IoT and robot systems for their operation to reduce human workload and increase profitability. However, the majority of the livestock and aquaculture industries in the Asian and Pacific countries are operated by small-scale farmers and fish folks. Introducing the same intelligent machines and systems which work for large-scale industry to those small-scale producers is not cost efficient and does not increase their income. Therefore, the development and adoption of smart technologies is becoming an important task so that small farmers, who are the basis of agriculture in Asian countries, can effectively respond to the above-mentioned challenges.

There are some experiences and research results on smart technologies from both the Livestock Research Institute (LRI) and Fisheries Research Institute (FRI) of the Council of Agriculture in Taiwan. The LRI has launched the Smart Farming (Leading Industry Technology Development and Application in Livestock Industry) project since 2017, evaluating the feasibility of introduction of foreign smart machinery and its self-development to simultaneously solve the problem of labor shortage and increase production on domestic dairies, and prioritizing the introduction of milking and feeding robots to alleviate labor shortage, increase milk yield and improve the quality of operation. In the aquaculture sector, using IoT and machine-to-machine (M2M) will promote the production efficiency of the aquaculture industry. Through both the fish growing recognition system and precise feeding system developed by the FRI, it could identify the ability of fish to take feed, body length of fish and decrease the feed cost. Increasing the water temperature and controlling the water quality of fishponds with solar power energy could save the power cost by NT$20,000 and 950 tons water per pond/per year.

This symposium will invite the experts and officers from FFTC’s member and partner countries in the Asian and Pacific region to share their experiences and expertise on the R&Ds and promotion of the adoption of intelligent livestock and aquaculture operations applicable for small-scale farmers in the region.

Currently, large scale livestock and aquaculture industries in the Asian and Pacific region have been actively introducing the IoT and robot systems for their operation to reduce human workload and increase profitability. However, the majority of the livestock and aquaculture industries in the Asian and Pacific countries are small scale farmers and fish folks. Introducing the same intelligent machines and systems which work for large scale industry to those small-scale producers is not cost efficient and does not increase their income. Therefore, the development and adoption of smart technology is becoming an important task so that small farmers, who are the basis of agriculture in Asian countries, can effectively respond to the above-mentioned challenges.

There are some experiences and research results on smart technologies from both the Livestock Research Institute (LRI) and Fisheries Research Institute (FRI) of the Council of Agriculture in Taiwan. The LRI has launched the Smart Farming (Leading Industry Technology Development and Application in Livestock Industry) project since 2017, evaluating the feasibility of introduction of foreign smart machinery and self-development of them to simultaneously solve the problem of labor shortage and increase production on domestic dairies, and prioritizing the introduction of milking and feeding robots to alleviate labor shortage, increase milk yield and improve the quality of operation. In the aquaculture sector, using IoT and machine-to-machine (M2M) will promote the production efficiency of the aquaculture industry. Through both the fish growing recognition system and precise feeding system developed by the FRI, it could identify the ability of fish to take feed, body length of fish and decreased the feed cost. To increase the water temperature and control the water quality of fishponds with solar power energy could save the power cost by NT$20,000 and 950 tons water per pond/per year.

This symposium will invite the experts and officers from FFTC’s member and partner countries in the Asian and Pacific region to share their experiences and expertise on the R&Ds and promotion of the adoption of intelligent livestock and aquaculture operations applicable for small-scale farmers in the region.

Objectives

• Do extension and promotion of intelligent operation for farmers and fish folks in the Asian and Pacific region.
• Exchange knowledge and experiences on livestock industry and aquaculture.
• Strengthen international cooperation on intelligent operations.
• Increase the income of farmers and fish folks of the livestock industry and aquaculture in the Asian and Pacific region.

Themes

• Session 1: Problems and solutions on small-scale livestock and aquaculture industries in the Asian and Pacific region—mapping out and promotion of advanced intelligent technologies.
• Session 2: Sharing on success cases in the application of advanced intelligent technologies in the livestock and aquaculture industries.
• Session 3: Advanced intelligent technologies from the private sector.
• Session 4 (General discussion): Policy measures for transformation to intelligent production: challenges and possible solutions.

Workshop highlights

Fourteen speakers from eight countries (Hungary, Indonesia, Japan, Korea, the Philippines, Taiwan, Thailand, and Vietnam) were invited to share their experiences and expertise on the R&Ds and promotion of the adoption of intelligent livestock and aquaculture operations applicable for small-scale farmers in the Asian and Pacific region. The presentation materials include 14 PPTs, 11 papers, and 6 videos. Key takeaways were summarized by presentation:

Keynote Session

Dr. Mei Ping Cheng, LRI, Taiwan (K1): Dr. Cheng emphasized the necessity of adopting Intelligent technology to overcome obstacles facing the livestock industry such as climate change and labor shortage. In Taiwan, dairy, swine, and poultry industries were the pilot items used to develop smart agriculture. The promotion routes of the research and development on smart agriculture includes smart production and digital service. This includes the poultry robot for image analysis, application of chicken voiceprint recognition system, smart waterfowl management system, the robotic milking system, and the smart pig industry information database. Cross-disciplinary technologies can enhance continuous innovation, transformation, and industrial upgrade. Livestock production based on the advanced cross-disciplinary technologies, such as information communication technology (ICT), Internet of Things (IoT), Big Data analysis, and Block Chain, combining with labor-saving mechanical equipment, assistive device, and sensors can not only reduce the labor demand, but also provide farmers with more efficient management models. 

Dr. Chung-Cheng Chang, NTOU, Taiwan (K2): Despite the continuous increase in world fishery production, the demand for aquatic products is increasing due to the increase in the world population. So, it is necessary to increase productivity through smart technology. The 3D smart marine farm system integrates AI technology, IoT technology, aquatic products breeding technology, automation technology and underwater technology to use omni AIoT aquaculture technology. Especially, it strengthens the integration of rotorcraft, remotely operating vehicle, unmanned boat, and central control room. It uses remote intelligent control and allows aquaculture practitioners to work at the base and central control room. This system is targeted to be applied to a large number of offshore cage culture farming. Using this system, operators can optimize the farming process according to the recommendations, and can obtain real-time farming information and analysis results, so as to minimize the farming cost and maximize the catch.

Session 1: Intelligent technology for livestock and aquaculture industry: implications, development, and promotion

Dr. Heru Sukoco, IPB, Indonesia (S1-1): Smallholder farming communities need to consolidate themselves, their livestock, and their productive assets to build a collective business. The smallholder farmers empowerment system (called SPPR) has been designed and implemented to consolidate the community of smallholder farmers as owners of 98% of the local livestock population with an ownership scale of only 2-3 cows per farmer to realize an area-based collective farming business. The application of digital technology and IoT can be easily applied to smallholder farmers by following the SPPR system. Through this, the business processes that occur between farmers and stakeholders such as consumers, government industry, and universities can be established quickly, accurately, with proper accountability, and have high-value benefits for all parties.

Dr. Jong-bok Kim, NIAS, Korea (S1-2): In Korea, many ICT devices and facilities are supplied to livestock farms. However, the technology has been focused on automation devices to reduce labor. It is called the 1^st generation smart farm. Meanwhile, livestock management is still dependent on farmer’s experiences. Now, it is time to change from experience-based livestock farming to data-driven livestock farming, which focuses more on productivity enhancement. It is called the 2^nd generation smart farm. For this, Korea is focusing on developing data acquisition technology and various data applications such as ICT device integrated management system, ICT device standardization, automatic bedding material spreader, biometric information monitoring devices, image-based livestock weight estimation. In the future, many data utilization models and services for livestock management will be provided to the farms, which will help farmers to have better productivity, better working environment, and better epidemic prevention environment.

Dr. Cao Le Quyen, VIFEP, Vietnam (S1-3): Smart technologies can be applied at 7 nodes within Vietnam’s shrimp value chain, including water environmental monitoring and response, shrimp behavior monitoring and auto-feeding devices, e-trading, e-commerce, e-traceability, e-finance and e-extension. However, smart technology and operations have just been initiated only at the pond water environmental monitoring and response, and shrimp behavior monitoring and auto-feeding devices with various challenges of high initial costs, less effective monitoring, unstable monitoring data, etc. To facilitate the adoption of intelligent technologies in local shrimp farming and the value chain, solutions such as baseline survey, R&D investment for durable sensors and cost-effective intelligent operation system, a policy to provide foundation for e-traceability and online management, and pilot initiative on e-trading, e-commerce, e-finance and e-extension are required.

Dr. Chona Camille V. Abeledo, DLSU, Philippines (S1-4): Dr. Abelado introduced a free mobile application called ‘Crabifier’, which helps mangrove crab farmers identify their favorite king mangrove crab species among the juveniles of three wild-caught mangrove crab species. The physical features that drive Crabifier were identified through image analysis and was validated through DNA barcoding. A convolutional neural network algorithm was created, using the identified features, to run the app. Since its deployment in 2019, Crabifier has been shown to help validate the identity of hatchery raised juveniles for sale to prevent any issues of fraud between traders and pond owners, with its 92.4% accuracy. The next step for the technology is to create an automated sorter for the thousands of crablets that need to be processed per day.

Session 2 – Success cases in the application of advanced intelligent technologies in the livestock and aquaculture industries

Professor Dr. Chaiyapoom Bunchasak, KU, Thailand (S2-1): Dr. Chaiyapoom introduced the Artificial Neural Networks (ANNs) which Kasetsart University has developed to predict the methionine requirement of broiler chickens (1−42 days of age). ANNs can also precisely predict the standardized ideal digestibility of lysine in full-fat soybean feed for pigs. Moreover, combinations of ANNs with image processing are being studied to improve the precision of animal health monitoring. An image segmentation and deep learning approach can be used to automate the process of scoring foot pad dermatitis based on images of chickens’ feet, which can help to minimize the subjective bias inherent in manual scoring. Recently, deep learning with a convolutional neural network was applied for disease diagnosis in shrimps.

Dr. Jeng-Bin Lin, LRI, Taiwan (S2-2):  Dr. Lin introduced several intelligent livestock farming technologies that has been developed in Taiwan. The robotic milking system can each serve 60 to 70 heads lactating cows per day and could save 12 hours of workload a day. The average daily milk production per head has increased to 33 kg, and the average milking frequency is about 2.9 times per head per day. The robotic feed pusher is a smart forage-feeding machine which has effectively replaced the manpower requirement of 3 to 4 hours per day, that could increase the eating times and intake of the cows. The milk yield has stably been increasing by 3 to 8%. Voice collecting system could distinguish the difference between the birthing cow’s moo or normal sounds and the accuracy rate could reach 90%. Other technologies such as Automatic Calf Feeder (ACF), Cleaning Robot for Calf Barn Floor (CRCBF), and Unmanned Aerial Vehicle (UAV) have also been proven effective in reducing feed and labor costs as well as in increasing productivity.

Dr. Yuan-Nan Chu, NTU, Taiwan (S2-3): Dr. Chu discussed the reason why the salmon industry moved on with the new technology while others didn’t. It turns out that the maturity of the intelligent technology for each sector and its cost effectiveness are the two main factors to consider in adopting this new tool. The turbidity of water in land-based ponds has hindered the development of vision-based intelligent applications. The much lower production value of land-based ponds restricts the investment of expensive hardware and software. A possible way to alleviate these problems is through intensification of the ponds. The intensification of land-based ponds could bring about a new wave of interest in intelligent technology in the future. Also, he introduced several intelligent shrimp farming systems which NTU is developing such as underwater video system, shrimp and feed recognition, and smart feeding.

Dr. Nobuo Ezaki, Toba College, Japan (S2-4): The timer-type automatic feeder used in red sea bream farming can only feed a fixed amount at a set time, which leads to overfeeding, and eventually to wasted feed costs, deterioration of water quality, and bottom sediment pollution. Dr. Ezaki introduces a technology realizing activity determination using classical image processing techniques and machine learning. When the fishes no longer eat during feeding, feeding is stopped and wasteful feeding is eliminated. The state of the fish being fed is captured by a camera, and the activity is determined using the variation in the brightness value when gray-scaled by the classical image processing technique as an index. Activity judgment is performed using a model of machine learning. During feeding, the white bubbles are produced by the snapper feeding on the surface; when the activity is low, the bubbles are no longer produced, and the mechanism use these differences to identify them.

Session 3 – Advanced intelligent technologies from the private sector

Mr. Jung-Hung Yen, Taijiang Agricultural Biotechnology CO., LTD., Taiwan (S3-1): Smart aquaculture in Taijing Agriculture Biotechnology Co. Ltd. is fully controllable through artificial systems. The feeds are mixed with compound formulas of various natural herbs which are adjusted accordingly when it is necessary. It can enhance the immunity and cold resistance of the fish and shrimps. Adopting “Lactic acid bacteria” and “Bacillus spp” from microbial agents, can effectively purify the water, reduce sediments, and eliminate the earthy smell. They developed a Psychic Fisherman App to monitor the water quality of fish farms using the image technique for smart aquaculture. In cooperation with NCKU, the image recognition technologies such as watercolor and waterwheel tail are being developed, so that they can make an image identification system of water quality. The company has adopted the AIOT smart farming system that integrates lighting, gas, water quality, nutrient solution/feeding, video, audio, energy management, and monitoring.

Mr. Steve Kim, uLikeKorea, Korea (S3-2): U-Like Korea is providing live care services based on orally injected sensors (bio capsules) for livestock. The Wi-Fi-based bio-capsule is a non-toxic material, has a battery life of 1 year (calf and sheep) to 5 years (cow), settles in rumen, transmits data 100 times a day, and real-time GPS tracking is possible. The live care service is provided through mobile and web, and provides services such as early disease detection, heat detection, optimal calving time prediction, and on-farm management. The livestock biodata accumulated in this process is used for biometric analysis based on deep learning to support disease control and labor management with 98% accuracy. The data driven livestock health management will allow farmers to focus more on the actual farm management, allowing the farms to grow and be more productive. It will also allow other businesses to grow such as data exchange among the farmers to share information on how best to deal with certain diseases, etc.

Mr. Tamás Szobolevszki, Bábolna Tetra Ltd., Hungary (S3-3): The Bábolna TETRA Ltd. is a privately owned poultry genetic center founded in 1967. Bábolna TETRA puts the latest scientific methods at the service of quality selection. It provides the suitable hybrids for all markets with the strength in the size of pedigree flocks and test capacity, constantly evolving genetic research, and feedbacks from partners and farmers around the word. The basis of its genetic work is data collection. It collects more than 70,000 individual data of relevant performance traits (egg production, egg quality, mortality) of each poultry per day.

Mr. Siang Bin Wu, Fang Yuan Farm, Taiwan (S3-4): Fang Yuan Farm has adopted 3 intelligent technologies which the Livestock Research Institute (LRI) has further developed. The ‘Intelligent waterfowl egg laying recognition system’ is used to identify the egg production of each breeding goose and then eliminate geese of low-egg production to improve the reproductive efficiency. The ‘intelligent system of environmentally controlled goose house’ monitors lighting, fan, water curtain wall and canvas lifting, access control, power, remote monitoring and anomaly reporting system, etc., and coordinates and integrates software, hardware interface and transmission information. The ‘intelligent system of feeding management’, which can simultaneously measure the body weight of the goose, uses 12 cameras to collect the image data of the feed intake, water play, drinking, and activity in the breeding geese farm, which will be used as the deep learning data for artificial intelligence (AI) image recognition of abnormal behavior of the breeding geese.

Session 4 (General Discussion) – Policy measures for transformation to intelligent production: challenges and possible solutions

(1) Policy for R&D promotion

Dr. Chaiyapoom Bunchasak, KU, Thailand: Currently, investments and efforts are actively being made to monitor environmental variables related to the livestock industry, detect problems early, and derive solutions. However, the current method, which mainly relies on human analysis, has limitations. The AI technology such as Artificial Neural Network (ANN) is expected to specifically present the optimal proposal considering various variables. Therefore, this should be one of the priorities of R&D directions.

(2) Policy for the transformation of intelligent production

Dr. Chung-Cheng Chang, NTOU, Taiwan: AIoT & Symbiosis of fisheries and electricity can attract young people to join the farming because it can improve the breeding environment, strengthen the quality of aquaculture, and increase aquaculture production. Therefore, the policy promoting integration of AIoT into the Symbiosis of fisheries and electricity is needed. Also, despite the increasing number of intelligent aquaculture equipment manufacturers, the degree of system intelligence and the quality stability are insufficient, and the price is not cheap enough. That’s why many aquaculture operators still use less intelligent equipment. It is urgent for the government to strengthen the encouragement of a policy to promote the production of several manufacturers with excellent business performance and international competitiveness.

Dr. Chaiyapoom Bunchasak, KU, Thailand: In Thailand, while chicken and swine production is commercialized by giant companies, the ruminant industry such as dairy farming is run by small holder farmers. Unlike giant businesses which have a lot of resources, the small holder farmers need a lot of technical support, but the reality is that it is difficult to transfer technology rapidly because the characteristics of each farm are different. This creates a major problem in transformation of intelligent production.

(3) Business models of intelligent systems

Dr. Mei-Ping Cheng, LRI, Taiwan: The Yuan Jin Chuang Smart Poultry Farmers Alliance handles all processes from hatching to breeding, slaughtering, processing, and marketing. By applying intelligent technology to all processes, the company can achieve high production efficiency and high quality, prevent mismatch between production and sales, and strengthen competitiveness. Small farmers can gain stability by participating in the breeding process. All breeding facilities are standardized, and the SOP of breeding process was established. Contract farmers can reduce costs and enhance biosecurity through real-time monitoring and automated intelligent control. The company analyzes customer feedback and sales trend data to determine the number of contract production. Through this, farmers can obtain maximum profits and can recoup their investment within three years.

Dr. Chona Camille V. Abeledo, DLSU, Philippines: The intelligent business model for aquaculture in the Philippines is a bit difficult. This is because fishfolks prefer their traditional methods rather than the use of intelligent technologies. Therefore, to accelerate the use of technology in the Philippines, it is necessary to strengthen communication and relationships with the local community.  Fish folks communities want to be involved in the technology development stage and do not want research institutes or large corporations to sell or show them the already developed technology. So, to drive this innovation in the Philippines, there is a need to work with the local communities that they want to help. In particular, the gap among business entities in the Philippines is very large, with some well-established companies and some very poor fishermen. Therefore, it is also very important to maintain a balance between tradition and innovation.

Suggestions and conclusions

• Put more emphasis on cross-disciplinary technologies that can enhance continuous innovation, transformation, and industrial upgrade.
• Lower the threshold for cage culture that requires high capital, technology and experience, and high risks through the integration of AIoT, automation technology, and underwater technology.
• Consolidate small holder farmers into collective entrepreneurs so that the application of digital technology and IoT can be easily applied.
• Move the focus from labor-saving automation to data-driven livestock farming, which focuses more on productivity enhancement.
• Conduct baseline survey, R&D investments for durable sensors and cost-effective intelligent operation systems. Facilitate policy to provide foundation for e-traceability and online management, and pilot initiative on e-business.
• Intensify the land-based ponds so that the capacity and necessity for more investments on the intelligent technologies which could help reduce risk and increase revenue.
• Strengthen cooperation between the private sector, universities, public research institutes, and the government for practical technology experiments and development.
• Put the priorities of R&D directions to the AI technology to specifically present the optimal proposal considering various variables.
• Establish the policy promoting integration of AIoT into the Symbiosis of fisheries and electricity and promote the production of several manufacturers with excellent business performance and international competitiveness.
• Strengthen communication with local community and balance tradition and innovation to enable small farmers to actively apply intelligent technologies.

Outcomes

71 people registered for the workshop, including participants from the Taiwan (70%), the Philippines (11%), Korea (6%), Thailand (4%), U.S.A (3%), Malaysia (3%), Cambodia, and Indonesia. Among the registrants, 48 were from the public sectors, 18 (about 25%) from the private sectors, and others from research institutes, universities, and international organizations. 51 people attended the hybrid workshop on-site. The workshop was livestreamed and broadcasted on two platforms, the Cisco Webex Event (Max. 1,000) and the FFTC Facebook pages. The Facebook video stream reached 144 views during the workshop. The Feedback form was circulated to the registered participants immediately after the workshop. Nearly all respondents were very satisfied with the workshop on all aspects (content, relevance and logistics). Overall, the workshop was regarded as successful in terms of planning, coordination, and execution.

Workshop videos can be watched at:
https://www.facebook.com/fftcforasiaandthepacific/videos

More information can be viewed on the workshop website:
https://km.fftc.org.tw/workshop/5