Review of Energy-Efficient Embedded System Acceleration of Convolution Neural Networks for Organic Weeding Robots

The vegetable crop cultivation in Polewali Mandar Regency has a great opportunity due to the high market demand for vegetable products. Environmental issues are a global concern for the competitiveness of agricultural products. The quality of organic agriculture or environmentally friendly product has become a concern for consumers regarding food safety, health, and long-life habituation . The farming community in Galung Lombok village, Tinambung District, Polewali Mandar Regency depends highly on the use of synthetic chemical fertilizers and pesticides. On the other hand, harvest waste, household waste and coconut water as a byproduct of coconut processing have not been utilised effectively. Thus, this community service program aimed to utilise household organic waste as organic fertiliser and utilise coconut water as plant growth hormone; as well as educate farmers in implementing environmentally-friendly agricultural businesses. The method used were counselling, training, discussion and demonstration to 25 farmers. The program was able to increase the motivation and knowledge of farmers to implement the cultivation of environmentally friendly vegetable crops by using organic fertilizer from household waste and ZPT from coconut water. The vegetable crop cultivation in Polewali Mandar Regency has a great opportunity due to the high market demand for vegetable products. Environmental issues are a global concern for the competitiveness of agricultural products. The quality of organic agriculture or environmentally friendly product has become a concern for consumers regarding food safety, health and long-life habituation . The farming community in Galung Lombok village, Tinambung District, Polewali Mandar Regency depends highly on the use of synthetic chemical fertilizers and pesticides. On the other hand, harvest waste, household waste and coconut water as a byproduct of coconut processing have not been utilised effectively. Thus, this community service program aimed to utilise household organic waste as organic fertiliser and utilise coconut water as plant growth hormone; as well as educate farmers in implementing environmentally-friendly agricultural businesses. The method used were counselling, training, discussion and demonstration to 25 farmers. The program was able to increase the motivation and knowledge of farmers to implement the cultivation of environmentally friendly vegetable crops by using organic fertilizer from household waste and ZPT from coconut water.

Biopesticides are perceived as a feasible alternative to chemical pesticides, providing an effective approach to pest management while mitigating the undesirable effects of chemical pesticide overuse. Yet, due to the distinctive attributes of the two types of pesticides, farmers often adopt a combination of both. This study aimed to probe the interdependent relationship underlying the mixed use of chemical pesticides and biopesticides by farmers in the rice–crayfish integrated system (RCIS) in Hubei province, China. We also sought to identify influencing factors and associated consequences of this practice. Given that the adoption of chemical pesticides and biopesticides by farmers is not mutually exclusive, a multivariable probit model was utilized to estimate simultaneous pesticide applications. Survey data from 736 RCIS farmers revealed that 10.50% of the sample farmers employ a combination of chemical pesticides and biopesticides. A substitution relationship was observed between the adoption of chemical pesticides and biopesticides by farmers, whereas no significant complementary relationship was found in the application of the two types of pesticides. The primary motivation for the mixed use of pesticides by farmers is to achieve superior pest control efficacy. Notably, when integrating the use of pesticides, farmers pay less attention to factors such as resource endowments and multiple production objectives. Further analysis indicated that farmers display significant concern for pesticide attributes including insecticidal efficacy, toxicity, insecticidal spectrum, and validity period. The mixed use of pesticides has resulted in a decreased frequency of pesticide application but has escalated the cost. While the mixed use of pesticides elevated the selling price of rice, no significant improvement was observed in yield and output value. The insights garnered from this study offer strategic implications for policymakers on how to enhance the judicious adoption of pesticides by farmers.

This research was aimed to analyze: behavior of farmers on chemical pesticides use in chilli crop farming, factors that influence the behavior of farmers in the use of chemical pesticides, and performance of chemical pesticides use by the chilli crops farmers. This research was conducted in Pekon Gisting Atas, Gisting District, Tanggamus Regency. Location of this research was chosen intentionally with the consideration that Pekon Gisting Atas has a number of chilli crop farmers and the circulation of chemical pesticides is more than in other pekon. Census method was used, and there were 33 chilli crop farmers as the research respondents. This research was conducted from October 2012 to May 2013. The methods used in this research were methods of case and direct observation. Descriptive and path analyses were used to analyze data. The results showed that behavior of farmers in the use of chemical pesticides in Pekon Gisting Atas was classified good enough. Factors that influenced the behavior of farmers in the use of pesticides were the farming experience, attitude, and farmers’ household income. The performance of chemical pesticide used by chilli crop farmers was as follows: insecticide, fungicide, bactericide, and herbicide was the types of pesticides widely used to control chilli plant pests and diseases; chemical pesticides was sprayed by farmers on scheduled time regardless of the presence or absence of pests and diseases. Key word: behavior of farmers, chilli, path analysis, pesticides, variability

Objective: To verify the risk factors associated with the use of agrochemicals, through the degree of compliancewith Mexican regulations on the Good Use and Management of Pesticides (BUMP).Design/Methodology/Approach: One-hundred producers, selected from a group of 300 producers who make up the register of the Sistema Producto Nopal in Morelos, Mexico, answered the questionnaire. The questionnaire included plot information, market data, and open questions about the compliance with the checklist for the recognition of areas where Good Use and Management of Pesticides (BUMP) standards are applied in the primary production of vegetables of the National Service of Health, Food Safety and Agro Food Quality (SENASICA)Results: Out of all the one-hundred producers surveyed, 57 were recognized in BUMP and 43 were not. Whether producers complied or not with these regulations and could have access to the export market depended on the crop area and the economic income it represents. The greatest non-compliance was caused by the non-authorized use of pesticides in the production of nopal, which represents a contamination risk for the worker, the environment, and the consumers.Study Limitations/Implications: Studies should be carried out about the impact on the health of farm workers and the environment resulting from the inappropriate use of pesticides.Findings/Conclusions: Although the implementation of BUMP in the nopal cultivation minimized the risks of contamination or poisoning, the non-authorized use of synthetic pesticides for its cultivation poses contamination risks to the worker, consumer, and environment. Therefore, strategies must be generated in the domestic market to guarantee the same food safety than in the international market.

Farmer's own decisions outweigh management strategies in reducing pesticide use in apple orchards

The haphazard and irrational use of chemical pesticides has led to pest resurgence and pesticide resistance, posing a significant challenge in the management of economically important pests. Farming systems built on integrated pest management (IPM) can minimize the use of chemical pesticides and secure soil, plant, and human health by controlling pests at the same time. Bactrocera cucurbitae is a serious pest of cucurbitaceous crops including more than other 70 hosts. Though various control strategies have been developed against melon fruit flies, fewer efforts have been paid to establish effective and efficient integrated pest management programs. This review provides an overview of the life cycle and behaviour of melon fruit fly as well as a detailed analysis of the various IPM strategies that have been developed for their control. These strategies include cultural practices, such as weeding, removal of alternate hosts, and field sanitation, biological control methods such as use of natural enemies, and entomopathogenic fungi (Beauveria bassiana, B. pseudobassiana), use of botanicals (neem, jholmol), and chemical control measures, such as the use of pesticides.

Convolution neural networks (CNNs) as one of today's main flavor of deep learning techniques dominate in various image recognition tasks. As the model size of modern CNNs continues to grow, neural network compression techniques have been proposed to prune the redundant neurons and synapses. However, prior techniques disconnect the software neural networks compression and hardware acceleration, which fail to balance multiple design parameters, including sparsity, performance, hardware area cost, and efficiency. More concretely, prior unstructured pruning techniques achieve high sparsity at the expense of extra performance overhead, while prior structured pruning techniques relying on strict sparse patterns lead to low sparsity and extra hardware cost. In this article, we propose OMNI, a framework for accelerating sparse CNNs on hardware accelerators. The innovation of OMNI stems from that it uses hardware amenable on-chip memory partition patterns to seamlessly engage the software CNN model compression and hardware CNN acceleration. To accelerate the compute-intensive convolution kernel, a promising hardware optimization approach is memory partition, which divides the original weight kernels into several groups so that the different hardware processing elements can simultaneously access the weight. We exploit the memory partition patterns including block, cyclic, or hybrid as a means of CNN compression patterns. Our software CNN model compression balances the sparsity across different groups and our hardware accelerator employs hardware parallelization coordinately with the sparse patterns, leading to a desirable compromise between sparsity and performance. We further develop performance models to help the designers to quickly identify the pattern factors subject to an area constraint. Last, we evaluate our design on application specific integrated circuit (ASIC) and field-programmable gate array (FPGA) platform. Experiments demonstrate that OMNI achieves 3.4×- 6.2× speedup for the modern CNNs, over a comparably ideal dense CNN accelerator. OMNI shows 114.7× energy efficiency improvement compared with GPU platform. OMNI is also evaluated on Xilinx ZC706 and ZCU102 FPGA platforms, achieving 41.5 GOP/s and 125.3 GOP/s, respectively.

Agricultural productivity is the key to global food security. Modern conventional farming has substantially increased food production, but at the expense of serious environmental harm [[1][1]]. By contrast, organic production is regarded as a suitable and more sustainable alternative owing to its

Recently neural network accelerators have grown into prominence with significant power and performance efficiency improvements over CPU and GPU. In this paper, we proposed two safety design techniques include Algorithm Based Atomic Error Checking-1 (ABAEC-1) and ABAEC-2 for a Weight Stationary (WS) Convolutional Neural Network (CNN) accelerator focusing on low latency and low overhead error detection and correction with no performance degradation. The proposed design techniques not only detect the errors on-the-fly but also perform error diagnosis to localize the errors to a Processing Element (PE) for on-line fault management and recovery. We applied the design techniques on an industry quality CNN accelerator and demonstrated that we could achieve the required Diagnostic Coverage (DC) goal with minimal area and power overhead for selected configurations. Furthermore, we discussed methods to extend the proposed techniques to other dataflow architecture.

GPUs became the reference platform for both training and inference phases of Convolutional Neural Networks (CNN) due to their tailored architecture to the CNN operators. However, GPUs are power-hungry architectures. A path to enable the deployment of CNNs in energy-constrained devices is adopting hardware accelerators for the inference phase. The design space exploration of CNNs using standard approaches, such as RTL, is limited due to their complexity. Thus, designers need frameworks enabling design space exploration that delivers accurate hardware estimation metrics to deploy CNNs. This work proposes a framework to explore CNNs design space, providing power, performance, and area (PPA) estimations. The heart of the framework is a system simulator. The system simulator front-end is TensorFlow, and the back-end is performance estimations obtained from the physical synthesis of hardware accelerators, not only from components like multipliers and adders. The first set of results evaluate the CNN accuracy using integer quantization, the accelerators PPA after physical synthesis, and the benefits of using a system simulator. These results allow a rich design space exploration, enabling selecting the best set of CNN parameters to meet the design constraints.

Concentrations, distribution, sources, and ecological risk of seven heavy metals including chromium (Cr), nickel (Ni), copper (Cu), zinc (Zn), lead (Pb), cadmium (Cd), and mercury (Hg) in agricultural topsoil samples of the Three Gorges Dam region, China were investigated in this study. Among seven heavy metals, Zn had the highest mean concentration (149 mg kg(-1)) in the agricultural topsoil, followed by Cr (66 mg kg(-1)), Cu (52.2 mg kg(-1)), Pb (13.0 mg kg(-1)), Ni (8.5 mg kg(-1)), Cd (0.29 mg kg(-1)), and Hg (0.08 mg kg(-1)). Enrichment factor (EF) values of Zn, Cu, Cd, and Hg were higher than 1.5, indicating that Zn, Cu, Cd, and Hg were the major pollutants in this study area. The average potential ecological risk index (RI) value was 147, suggesting that heavy metals in the agricultural topsoil in the study area had a low ecological risk. The result of factor analysis (FA) and correlation analysis showed that long-term use of chemical fertilizer and pesticides, natural rock weathering, and atmospheric deposition were the several main sources of seven heavy metals in agricultural topsoil of the Three Gorges Dam region. Factor analysis-multiple linear regression (FA-MLR) results indicated that the most important source in this area was long-term use of chemical fertilizer and pesticides, which contributed 70 % for Cu and Zn, 62 % for Cd, and 72 % for Hg. More attention must be paid to the extensive use of chemical fertilizers and pesticides containing heavy metals which have been accumulated in the agricultural soil.

Deli Serdang is an agricultural center in North Sumatra where farming activities are still done conventionally using chemical fertilizers and synthetic pesticides. The frequent use of synthetic pesticides continuously leads to negative impacts such as the emergence of pest resistance and resurgence. Because the pests have become resistant, farmers will continue to use synthetic pesticides by increasing the dosage and application time so that production can be achieved. The purpose of this research is to understand the actual impact, identify sustainable solutions, and promote agriculture that is safer, environmentally friendly, and sustainable. By analyzing the impact of pesticide use in agriculture considering various aspects. The method used is qualitative descriptive through interviews and observations. The excessive use of pesticides in agricultural ecosystems pollutes the soil and water and kills beneficial non-target organisms, which can disrupt the ecosystem balance and reduce agricultural biodiversity. Meanwhile, the impact on the health of farmers includes health risks, including acute and chronic poisoning. Farmers and agricultural workers may experience respiratory problems, skin irritation, nervous system damage, and long-term risks such as cancer.

The need to implement an agricultural system taking into account sustainable development has fostered many initiatives to develop alternative methods in order to reduce the use of chemical synthetic pesticides for plant protection. Among these alternatives, the use of Biocontrol Agents (BCAs) or biopesticides has aroused increasing interest because of their ecological advantages. Their commercialization now increases but the part they occupy in Plant Protection Products (PPP) market is still marginal. After defining what are the BCAs (biopesticide) products, this chapter analyses the factors that encourage or hamper this development and also considers the room of BCAs (biopesticides) and the trends in the global PPP market.

The efficacy of convolutional neural networks (CNNs) has led to their extensive application in picture recognition and categorization. Nevertheless, embedded systems face challenges when it comes to storing the massive volumes of weight data required by CNNs in their on-chip memory. Pruning a convolutional neural network (CNN) model can reduce its size with no impact on accuracy; however, when used with a parallel architecture, the resulting model will be slower to execute. A CNN compression method that is hardware-centric is introduced in this paper. A model of a deep neural network (DNN) will include "pruning layers (P-layers)" and "no-pruning layers (NP-layers)" attached to it. For effective and parallel processing, an NP-layer employs a regular distribution for its weights. Pruning causes a P-layer to have an irregular form, yet the high compression ratio it yields makes the shape undesirable. Using uniform and gradual quantization methods, we can achieve a processing efficiency/compression ratio trade-off with a small loss of precision. Integrating multiple parallel finite impulse response (FIR) filters into an NP-layer distributed convolutional architecture, we further improve the regular model. The P-layer irregular sparse model is suggested to use an ADF-based processing element. The suggested compression strategy and hardware architecture can be utilised by a hardware/algorithm co-optimization (HACO) method to run an NP→P hybrid compressed CNN model on FPGAs. The VGG-16 network achieves a compression ratio of 27.5× with a top-5 accuracy loss of 0.44% by using a hardware accelerator on a single FPGA chip without off-chip memory. A result of 83.0 FPS, or 1.8 times quicker, was achieved when the compressed VGG-16 model was implemented on a Xilinx VCU118 evaluation board for image applications. Index Terms - CNN, Accelerators, Co-Optimization.

BackgroundNew bioinsecticides with novel modes of action are urgently needed to minimise the environmental and safety hazards associated with the use of synthetic chemical pesticides and to combat growing levels of pesticide resistance. The pea seed albumin PA1b knottin peptide is the only known proteinaceous inhibitor of insect vacuolar adenosine triphosphatase (V-ATPase) rotary proton pumps. Oral toxicity towards insect pests and an absence of activity towards mammals makes Pa1b an attractive candidate for development as a bioinsecticide. The purpose of this study was to investigate if Pichia pastoris could be used to express a functional PA1b peptide and if it’s insecticidal activity could be enhanced via engineering to produce a fusion protein comprising the pea albumin protein fused to the mannose-specific snowdrop lectin (Galanthus nivalis agglutinin; GNA).ResultsWe report the production of a recombinant full-length pea albumin protein (designated PAF) and a fusion protein (PAF/GNA) comprised of PAF fused to the N-terminus of GNA in the yeast Pichia pastoris. PAF was orally toxic to pea (Acyrthosiphon pisum) and peach potato (Myzus persicae) aphids with respective, Day 5 LC50 values of 54 µM and 105 µM derived from dose–response assays. PAF/GNA was significantly more orally toxic as compared to PAF, with LC50 values tenfold (5 µM) and 3.3-fold (32 µM) lower for pea and peach potato aphids, respectively. By contrast, no phenotypic effects were observed for worker bumble bees (Bombus terristrus) fed PAF, GNA or PAF/GNA in acute toxicity assays. Confocal microscopy of pea aphid guts after pulse-chase feeding fluorescently labelled proteins provides evidence that enhanced efficacy of the fusion protein is attributable to localisation and retention of PAF/GNA to the gut epithelium. In contact assays the fusion protein was also found to be significantly more toxic towards A. pisum as compared to PAF, GNA or a combination of the two proteins.ConclusionsOur results suggest that GNA mediated binding to V-type ATPase pumps acts to potentiate the oral and contact aphicidal activity of PAF. This work highlights potential for the future commercial development of plant protein-based bioinsecticides that offer enhanced target specificity as compared to chemical pesticides, and compatibility with integrated pest management strategies.