Predicting the performance of anaerobic digestion using machine learning algorithms and genomic data

Deep-learning approach for developing bilayered electromagnetic interference shielding composite aerogels based on multimodal data fusion neural networks

A non-experimental approach to developing high-performance EMI shielding materials is urgently needed to reduce costs and manpower. In this investigation, a multimodal data fusion neural network model is proposed to predict the EMI shielding performances of silver-modified four-pronged zinc oxide/waterborne polyurethane/barium ferrite (Ag@F-ZnO/WPU/BF) aerogels. First, 16 Ag@F-ZnO/WPU/BF samples with varying Ag@F-ZnO and BF contents were successfully prepared using the pre-casting and directional freezing techniques. The experimental results demonstrate that these aerogels perform well in terms of averaged EMI shielding effectiveness (SET) up to 78.6 dB and absorption coefficient as high as 0.96. On the basis of composite ingredients and microstructural images, the established multimodal neural network model can effectively predict the EMI shielding performances of Ag@F-ZnO/WPU/BF aerogels. Notably, the multimodal model of fully connected neural network (FCNN) and residual neural network (ResNet) utilizing GatedFusion method yields the best root mean squared error (RMSE) and mean absolute error (MAE) values of 0.7626 and 0.4918, respectively, and correlation coefficient (R) of 0.9885. In addition, this multimodal model successfully predicts the EMI performances of four new aerogels with an average error of less than 5 %, demonstrating its strong generalization capability. The accuracy and efficiency of material property prediction based on multimodal neural network model are largely improved by integrating multiple data sources, offering new possibility for reducing experimental burdens, accelerating the development of new materials, and gaining a deeper understanding of material mechanisms.

Exploring the function of key species in different composting stages for effective waste biotransformation

Composting is a microbial-driven process that plays a vital role in recycling waste and promoting sustainable production. To develop more effective bioaugmentation strategies, this study examined three successive stages in an aerobic composting system, focusing on microbial community adaptation to high-temperature stress (mode_2) and nutrient-poor conditions (mode_3). The results revealed a shift from an r-strategy (rapid growth) to a K-strategy (thriving under resource-limited conditions). Community succession was predominantly driven by deterministic processes (>90 %) and exhibited strong cooperative interactions. Using multiple statistical approaches, key species were identified for each condition. These species enhanced microbial network connectivity under environmental stresses, increasing network edges by 29 %-35 %. Under high-temperature stress, Bacillus and Ureibacillus maintained core functions, while Chelativorans and Aeribacillus contributed to key metabolic pathways, including amino acid metabolism. In nutrient-poor conditions, Saccharomonospora and Pseudoxanthomonas enhanced overall system functionality, and Novibacillus played a key role in carbon and nitrogen cycling, particularly nitrogen fixation. Predictive models for microbial community stability (R2 = 0.68-0.97) were developed based on these key species to enable rapid assessment of system stability. Overall, this study identifies essential microbes involved in composting across different environmental conditions and clarifies their functional roles, providing valuable insights for optimizing aerobic composting efficiency and advancing waste resource management.

Enhancing anaerobic digestion of food waste with chemically vapor-deposited biochar: Effective enrichment of Methanosarcina and hydrogenotrophic methanogens

The effects and mechanisms of chemically vapor-deposited (CVD) biochar on anaerobic digestion (AD) remain unexplored. This study proposes a novel approach to simultaneously address plastic waste management and bioenergy production by applying CVD biochar in the anaerobic digestion of food waste. Results indicate that CVD biochar, particularly PE900, significantly reduces the lag phase (from 15 to 9 days) and increases methane yield by 46 % compared to the control. The effectiveness of polyethylene-derived biochar is further confirmed through three consecutive fermentation batches, where it consistently improves methane production. CVD biochar also alters extracellular polymeric substance (EPS) composition and enriches key microbial communities, including hydrogenotrophic methanogens and Methanosarcina. The nanofiber structure and higher sp2-hybridized carbon content of PE900 biochar are likely the main factors influencing EPS, microbial composition, and methane production performance. This study demonstrates the potential of CVD biochar for enhancing anaerobic digestion and valorizing plastic waste in waste-to-energy conversion.

A machine learning approach to feature selection and uncertainty analysis for biogas production in wastewater treatment plants

The growing demand for efficient waste management solutions and renewable energy sources has driven research into predicting biogas production at wastewater treatment plants. This study outlines a methodology starting with data collection from a full-scale plant, followed by detailed analysis to resolve potential issues. A notable advancement is the use of a robust machine learning model, fine-tuned with advanced optimization techniques. To enhance its utility, prediction intervals were incorporated to quantify uncertainty, providing decision-makers with reliable insights. Results revealed that the developed model performed well, explaining 82% of the variability in test data and delivering predictions closely aligned with actual biogas production. This reliability empowers more confident decision-making in wastewater treatment operations. The study also identified key factors influencing biogas output, categorizing them intosludge characteristics, operational practices, and sludge quantity. By focusing on most important adjustable parameters, operators can optimize processes and significantly improve biogas yields. This predictive capability, combined with an understanding of influencing factors and quantified reliability, offers notable advantages. It enables operators to enhance biogas production while providing decision-makers with reliable predictions to guide policy and resource management. These developments contribute to more sustainable and efficient waste management practices.

Machine learning predictions of high-strength RCA concrete utilizing chemically activated fly ash and nano-silica

This study explores the potential of RCA combined with nano silica and chemically activated fly ash to produce sustainable and high strength concrete. The research addresses the challenges posed by RCA's inferior mechanical and durability properties by incorporating SCM. A comprehensive experimental program includes 420 and 240 samples for compressive strength and acid resistance. Machine learning algorithms such as Decision Trees, Random Forest, XG-Boost, and Ada Boost are used to predict RCA concrete performance metrics, with XG-Boost achieve the highest predictive accuracy (R2 = 0.995) for compressive strength while random forest performance is better for acid resistance (R2 = 0.909). The findings demonstrate substantial improvement in mechanical performance and durability, under scoring the effectiveness of SCMs in optimizing RCA- based concrete. The integration of machine learning provides a robust framework for performance predictions, contributing to the advancement of sustainable and resilient construction materials.

Application of Machine Learning for FOS/TAC Soft Sensing in Bio-Electrochemical Anaerobic Digestion

This study explores the application of various machine learning (ML) models for the real-time prediction of the FOS/TAC ratio in microbial electrolysis cell anaerobic digestion (MEC-AD) systems using data collected during a 160-day trial treating brewery wastewater. This study investigated models including decision trees, XGBoost, support vector regression, a variant of support vector machine (SVM), and artificial neural networks (ANNs) for their effectiveness in the soft sensing of system stability. The ANNs demonstrated superior performance, achieving an explained variance of 0.77, and were further evaluated through an out-of-fold ensemble approach to assess the selected model's performance across the complete dataset. This work underscores the critical role of ML in enhancing the operational efficiency and stability of bio-electrochemical systems (BES), contributing significantly to cost-effective environmental management. The findings suggest that ML not only aids in maintaining the health of microbial communities, which is essential for biogas production, but also helps to reduce the risks associated with system instability.

Establishment of a Daqu Grade Classification Model Based on Computer Vision and Machine Learning

The grade of Daqu significantly influences the quality of Baijiu. To address the issues of high subjectivity, substantial labor costs, and low detection efficiency in Daqu grade evaluation, this study focused on light-flavor Daqu and proposed a two-layer classification structure model based on computer vision and machine learning. Target images were extracted using three image segmentation methods: threshold segmentation, morphological fusion, and K-means clustering. Feature factors were selected through methods including mean decrease accuracy based on random forest (RF-MDA), recursive feature elimination (RFE), LASSO regression, and ridge regression. The Daqu grade evaluation model was constructed using support vector machine (SVM), logistic regression (LR), random forest (RF), k-nearest neighbor (KNN), and a stacking model. The results indicated the following: (1) In terms of image segmentation performance, the morphological fusion method achieved an accuracy, precision, recall, F1-score, and AUC of 96.67%, 95.00%, 95.00%, 0.95, and 0.96, respectively. (2) For the classification of Daqu-P, Daqu-F, and Daqu-S, RF models performed best, achieving an accuracy, precision, recall, F1-score, and AUC of 96.67%, 97.50%, 97.50%, 0.97, and 0.99, respectively. (3) In distinguishing Daqu-P from Daqu-F, the combination of the RF-MDA method and the stacking model demonstrated the best performance, with an accuracy, precision, recall, F1-score, and AUC of 90.00%, 94.44%, 85.00%, 0.89, and 0.95, respectively. This study provides theoretical and technical support for efficient and objective Daqu grade evaluation.

Enhancing Genomic Prediction Accuracy of Reproduction Traits in Rongchang Pigs Through Machine Learning

The increasing volume of genome sequencing data presents challenges for traditional genome-wide prediction methods in handling large datasets. Machine learning (ML) techniques, which can process high-dimensional data, offer promising solutions. This study aimed to find a genome-wide prediction method for local pig breeds, using 10 datasets with varying SNP densities derived from imputed sequencing data of 515 Rongchang pigs and the Pig QTL database. Three reproduction traits-litter weight, total number of piglets born, and number of piglets born alive-were predicted using six traditional methods and five ML methods, including kernel ridge regression, random forest, Gradient Boosting Decision Tree (GBDT), Light Gradient Boosting Machine, and Adaboost. The methods' efficacy was evaluated using fivefold cross-validation and independent tests. The predictive performance of both traditional and ML methods initially increased with SNP density, peaking at 800-900 k SNPs. ML methods outperformed traditional ones, showing improvements of 0.4-4.1%. The integration of GWAS and the Pig QTL database enhanced ML robustness. ML models exhibited superior generalizability, with high correlation coefficients (0.935-0.998) between cross-validation and independent test results. GBDT and random forest showed high computational efficiency, making them promising methods for genomic prediction in livestock breeding.

Enhancing resource recovery from acid whey through chitosan-based pretreatment and machine learning optimization

Acid whey, a dairy byproduct with low pH and high organic content, presents disposal challenges but also potential for resource recovery. In this study, chitosan gel was synthesized and evaluated for turbidity reduction of acid whey. Machine learning (ML) models were employed to predict and optimize the pretreatment process, with the Random Forest algorithm achieving a prediction accuracy of 0.78. Using the Simulated Annealing algorithm, optimal conditions were identified, applying a 2.2 % chitosan solution gel at a dosage of 24 g/L to acid whey at pH 4.6 for 12 h, achieving a 91 % turbidity reduction, a significant improvement over the 71 % obtained prior to optimization. Validation experiments confirmed its effectiveness in predicting and optimizing the pretreatment process. These findings highlight the feasibility of ML in optimizing chitosan pretreatment and demonstrate chitosan gel as a cost-effective, efficient option for acid whey, with potential to enhance resource recovery in the dairy industry.

Integration of MALDI-TOF MS and machine learning to classify enterococci: A comparative analysis of supervised learning algorithms for species prediction

This research focused on distinguishing distinct matrix assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) spectral signatures of three Enterococcus species. We evaluated and compared the predictive performance of four supervised machine learning algorithms, K-nearest neighbor (KNN), support vector machine (SVM), and random forest (RF), to accurately classify Enterococcus species. This study involved a comprehensive dataset of 410 strains, generating 1640 individual spectra through on-plate and off-plate protein extraction methods. Although the commercial database correctly identified 76.9% of the strains, machine learning classifiers demonstrated superior performance (accuracy 0.991). In the RF model, top informative peaks played a significant role in the classification. Whole-genome sequencing showed that the most informative peaks are biomarkers connected to proteins, which are essential for understanding bacterial classification and evolution. The integration of MALDI-TOF MS and machine learning provides a rapid and accurate method for identifying Enterococcus species, improving healthcare and food safety.

Explainable machine learning-driven predictive performance and process parameter optimization for caproic acid production

In this study, four machine learning (ML) prediction models were developed to predict and optimize the production performance of caproic acid based on substrates, products, and process parameters. The XGBoost outperformed others, with a high R2 of 0.998 on the training set and 0.885 on the test set. Feature importance analysis revealed hydraulic retention time (HRT) and butyric acid concentration are decisive. The SHAP method offered profound insights into the interplay and cumulative effects of substrate composition, identified the synergistic effects between butyric acid and lactic acid, and emphasized adding glucose can benefit caproic with lactic acid co-fermentation. By integrating the Adaptive Variation Particle Swarm Optimization (AVPSO) algorithm, the optimal process conditions to achieve a maximum caproic acid production of 8.64 g/L was obtained. This study not only advances caproic acid production but contributes a versatile ML-driven strategy applicable to bioprocess optimizations, potentially transformative for sustainable and economically viable bioproduction.

Increasing methane production in an anaerobic membrane bioreactor for treating landfill leachate: Impact of organic concentration and HRT

The anaerobic biological treatment of landfill leachate frequently encounters the souring problems because of the high concentration of organic in landfill leachate. Nonetheless, the performance of anaerobic membrane bioreactor (AnMBR) is commendable in terms of removal of organic compounds. Hence, this study explored the effect of organic concentration and hydraulic retention time(HRT) on the removal performance of actual landfill leachate, additionally, carbon conversion through carbon mass balance analysis was analyzed, in order to determine the optimal treatment potential of AnMBR in treating landfill leachate. For HRT values between 14.5 h and 34.6 h, and the influent COD (Chemical Oxygen Demand) range of 12,773.33-15706.67 mg/L, AnMBR could efficiently treat landfill leachate. As HRT was fixed at 14.5 h and influent COD was around 12,206.7-15,373.33 mg/L, AnMBR achieved a maximum organic removal rate of 18.22 ± 0.51 kg COD/(m3∙d) with methane yield of 0.24 ± 0.01 m3 CH4/kg COD and methane content of 88.26%. Based on carbon mass balance, increasing COD concentration in the influent (less than 16,000 mg/L) boosted the conversion of organic compounds (45.19 ± 4.24%) into CH4; while decreasing HRT (more than 27.0 h) also promoted the conversion of organic compounds into CH4 (38.36-60.93%) resulting in a decreased TOC (Total Organic Carbon) loss by 2.02-7.19% with outflow. AnMBR may efficiently produce methane while treating landfill leachate by assessing the random forest model (RF) and adjusting the balance between HRT and influent COD concentration.

Plasticizers inhibit food waste anaerobic digestion performance by affecting microbial succession and metabolism

The widely existed plastic additives plasticizers in organic wastes possibly pose negative influences on anaerobic digestion (AD) performance, the direct evidence about the effects of plasticizers on AD performance is still lacking. This study evaluated the influencing mechanism of two typical plasticizers bisphenol A (BPA) and dioctyl phthalate on the whole AD process. Results indicated that plasticizers addition inhibited methane production, and the inhibiting effects were reinforced with the increase of concentration. By contrast, 50 mg/L BPA exhibited the strongest inhibition on methane production. Physicochemical analysis showed plasticizers inhibited the metabolism efficiency of soluble polysaccharide and volatile fatty acids. Microbial communities analyses suggested that plasticizers inhibited the direct interspecies electron transfer participators of methanogenic archaea (especially Methanosarcina) and syntrophic bacteria. Furthermore, plasticizers inhibited the methane metabolisms, key coenzymes (CoB, CoM, CoF420 and methanofuran) biosynthesis and the metabolisms of major organic matters. This study shed light on the effects of plasticizers on AD performance and provided new insights for assessing the influences of plasticizers or plastic additives on the disposal of organic wastes.

A data-driven approach for revealing the linkages between differences in electrochemical properties of biochar during anaerobic digestion using automated machine learning

Biochar is commonly used to enhance the anaerobic digestion of organic waste solids and wastewater, due to its electrochemical properties, which intensify the electron transfer of microorganisms attached to its large surface area. However, it is difficult to create biochar with both high conductivity and high capacitance, which makes selecting the right biochar for engineering applications challenging. To address this issue, two Auto algorithms (TPOT and H2O) were applied to model the effects of different biochar properties on anaerobic digestion processes. The results showed that the gradient boosting machine had the highest predictive accuracy (R2 = 0.96). Feature importance analysis showed that feedstock concentration, digestion time, capacitance, and conductivity of biochar were the main factors affecting methane yield. According to the two-dimensional (2D) partial dependence plots, high-capacitance biochar (0.27-0.29 V·mA) is favorable for substrates with low-solid content (< 19.6 TS%), while the high-conductivity biochar (80.82-170.58 mS/cm) is suitable for high-solids substrates (> 20.1 TS%). The software, based on the optimal model, can be used to obtain the ideal range of biochar for AD trials, aiding researchers in practical applications prior to implementation.

Decoding the Structural Diversity: A New Horizon in Antimicrobial Prospecting and Mechanistic Investigation

The escalating crisis of antimicrobial resistance (AMR) underscores the urgent need for novel antimicrobials. One promising strategy is the exploration of structural diversity, as diverse structures can lead to diverse biological activities and mechanisms of action. This review delves into the role of structural diversity in antimicrobial discovery, highlighting its influence on factors such as target selectivity, binding affinity, pharmacokinetic properties, and the ability to overcome resistance mechanisms. We discuss various approaches for exploring structural diversity, including combinatorial chemistry, diversity-oriented synthesis, and natural product screening, and provide an overview of the common mechanisms of action of antimicrobials. We also describe techniques for investigating these mechanisms, such as genomics, proteomics, and structural biology. Despite significant progress, several challenges remain, including the synthesis of diverse compound libraries, the identification of active compounds, the elucidation of complex mechanisms of action, the emergence of AMR, and the translation of laboratory discoveries to clinical applications. However, emerging trends and technologies, such as artificial intelligence, high-throughput screening, next-generation sequencing, and open-source drug discovery, offer new avenues to overcome these challenges. Looking ahead, we envisage an exciting future for structural diversity-oriented antimicrobial discovery, with opportunities for expanding the chemical space, harnessing the power of nature, deepening our understanding of mechanisms of action, and moving toward personalized medicine and collaborative drug discovery. As we face the continued challenge of AMR, the exploration of structural diversity will be crucial in our search for new and effective antimicrobials.

Adaptive prediction for effluent quality of wastewater treatment plant: Improvement with a dual-stage attention-based LSTM network

The accurate effluent prediction plays a crucial role in providing early warning for abnormal effluent and achieving the adjustment of feedforward control parameters during wastewater treatment. This study applied a dual-staged attention mechanism based on long short-term memory network (DA-LSTM) to improve the accuracy of effluent quality prediction. The results showed that input attention (IA) and temporal attention (TA) significantly enhanced the prediction performance of LSTM. Specially, IA could adaptively adjust feature weights to enhance the robustness against input noise, with R2 increased by 13.18%. To promote its long-term memory ability, TA was used to increase the memory span from 96 h to 168 h. Compared to a single LSTM model, the DA-LSTM model showed an improvement in prediction accuracy by 5.10%, 2.11%, 14.47% for COD, TP, and TN. Additionally, DA-LSTM demonstrated excellent generalization performance in new scenarios, with the R2 values for COD, TP, and TN increasing by 22.67%, 20.06%, and 17.14% respectively, while the MAPE values decreased by 56.46%, 63.08%, and 42.79%. In conclusion, the DA-LSTM model demonstrated excellent prediction performance and generalization ability due to its advantages of feature-adaptive weighting and long-term memory focusing. This has forward-looking significance for achieving efficient early warning of abnormal operating conditions and timely management of control parameters.

Machine learning for high solid anaerobic digestion: Performance prediction and optimization

Biogas production through anaerobic digestion (AD) is one of the complex non-linear biological processes, wherein understanding its dynamics plays a crucial role towards process control and optimization. In this work, a machine learning based biogas predictive model was developed for high solid systems using algorithms, including SVM, ET, DT, GPR, and KNN and two different datasets (Dataset-1:10, Dataset-2:5 inputs). Support Vector Machine had the highest accuracy (R2) of all the algorithms at 91 % (Dataset-1) and 87 % (Dataset-2), respectively. The statistical analysis showed that there was no significant difference (p = 0.377) across the datasets, wherein with less inputs, accurate results could be predicted. In case of biogas yield, the critical factors which affect the model predictions include loading rate and retention time. The developed high solid machine learning model shows the possibility of integrating Artificial Intelligence to optimize and control AD process, thus contributing to a generic model for enhancing the overall performance of the biogas plant.

Correlations and impact of anaerobic digestion operating parameters on the start-up duration: Database construction for robust start-up guidelines

Anaerobic digestion (AD) has become a popular technique for organic waste management while offering economic and environmental advantages. As AD becomes increasingly prevalent worldwide, research efforts are primarily focused on optimizing its processes. During the operation of AD systems, the occurrence of unstable events is inevitable. So far, numerous conclusions have been drawn from full and lab-scale studies regarding the driving factors of start-up perturbations. However, the lack of standardized practices reported in start-up studies raises concerns about the comparability and reliability of obtained data. This study aims to develop a knowledge database and investigate the possibility of applying machine learning techniques on experimentation-extracted data to assist start-up planning and monitoring. Thus, a standardized database referencing 75 cases of start-up of one-stage wet continuously-stirred tank reactors (CSTR) processing agricultural, industrial, or municipal organic effluent in mono-digestion from 31 studies was constructed. 10 % of the total observations included in this database concern failed start-up experiments. Then, correlations between the parameters and their impacts on the start-up duration were studied using multivariate analysis and a model-based ranking methodology. Insights into trends of choices were highlighted through the correlation analysis of the database. As such, scenarios favoring short start-up duration were found to involve relatively low retention times (average initial and final hydraulic retention times, (HRTi) and (HRTf) of 26.25 and 20.6 days, respectively), high mean organic loading rates (average OLRmean of 5.24 g VS·d-1·L -1) and the processing of highly fermentable substrates (average feed volatile solids (VSfeed) of 81.35 g L-1). The model-based ranking of AD parameters demonstrated that the HRTf, the VSfeed, and the target temperature (Tf) have the strongest impact on the start-up duration, receiving the highest relative scores among the evaluated AD parameters. The database could serve as a reference for comparison purposes of future start-up studies allowing the identification of factors that should be closely controlled.

Applying machine learning to anaerobic fermentation of waste sludge using two targeted modeling strategies

Anaerobic fermentation is an effective method to harvest volatile fatty acids (VFAs) from waste activated sludge (WAS). Accurately predicting and optimizing VFAs production is crucial for anaerobic fermentation engineering. In this study, we developed machine learning models using two innovative strategies to precisely predict the daily yield of VFAs in a laboratory anaerobic fermenter. Strategy-1 focuses on model interpretability to comprehend the influence of variables of interest on VFAs production, while Strategy-2 takes into account the cost of variable acquisition, making it more suitable for practical applications in prediction and optimization. The results showed that Support Vector Regression emerged as the most effective model in this study, with testing R2 values of 0.949 and 0.939 for the two strategies, respectively. We conducted feature importance analysis to identify the critical factors that influence VFAs production. Detailed explanations were provided using partial dependence plots and Shepley Additive Explanations analyses. To optimize VFAs production, we integrated the developed model with optimization algorithms, resulting in a maximum yield of 2997.282 mg/L. This value was 45.2 % higher than the average VFAs level in the operated fermenter. Our study offers valuable insights for predicting and optimizing VFAs production in sludge anaerobic fermentation, and it facilitates engineering practice in VFAs harvesting from WAS.

Multi-omics reveals the mechanism of rumen microbiome and its metabolome together with host metabolome participating in the regulation of milk production traits in dairy buffaloes

Recently, it has been discovered that certain dairy buffaloes can produce higher milk yield and milk fat yield under the same feeding management conditions, which is a potential new trait. It is unknown to what extent, the rumen microbiome and its metabolites, as well as the host metabolism, contribute to milk yield and milk fat yield. Therefore, we will analyze the rumen microbiome and host-level potential regulatory mechanisms on milk yield and milk fat yield through rumen metagenomics, rumen metabolomics, and serum metabolomics experiments. Microbial metagenomics analysis revealed a significantly higher abundance of several species in the rumen of high-yield dairy buffaloes, which mainly belonged to genera, such as Prevotella, Butyrivibrio, Barnesiella, Lachnospiraceae, Ruminococcus, and Bacteroides. These species contribute to the degradation of diets and improve functions related to fatty acid biosynthesis and lipid metabolism. Furthermore, the rumen of high-yield dairy buffaloes exhibited a lower abundance of methanogenic bacteria and functions, which may produce less methane. Rumen metabolome analysis showed that high-yield dairy buffaloes had significantly higher concentrations of metabolites, including lipids, carbohydrates, and organic acids, as well as volatile fatty acids (VFAs), such as acetic acid and butyric acid. Meanwhile, several Prevotella, Butyrivibrio, Barnesiella, and Bacteroides species were significantly positively correlated with these metabolites. Serum metabolome analysis showed that high-yield dairy buffaloes had significantly higher concentrations of metabolites, mainly lipids and organic acids. Meanwhile, several Prevotella, Bacteroides, Barnesiella, Ruminococcus, and Butyrivibrio species were significantly positively correlated with these metabolites. The combined analysis showed that several species were present, including Prevotella.sp.CAG1031, Prevotella.sp.HUN102, Prevotella.sp.KHD1, Prevotella.phocaeensis, Butyrivibrio.sp.AE3009, Barnesiella.sp.An22, Bacteroides.sp.CAG927, and Bacteroidales.bacterium.52-46, which may play a crucial role in rumen and host lipid metabolism, contributing to milk yield and milk fat yield. The "omics-explainability" analysis revealed that the rumen microbial composition, functions, metabolites, and serum metabolites contributed 34.04, 47.13, 39.09, and 50.14%, respectively, to milk yield and milk fat yield. These findings demonstrate how the rumen microbiota and host jointly affect milk production traits in dairy buffaloes. This information is essential for developing targeted feeding management strategies to improve the quality and yield of buffalo milk.

Response of nutrients removal efficiency, enzyme activities and microbial community to current and voltage in a bio-electrical anammox system

The effects of different current intensities and voltage levels on nutrient removal performance and microbial community evolution in a Bio-Electrical Anammox (BEA) membrane bioreactor (MBR) were evaluated. The nitrogen removal efficiency increased with the current intensity within the range of 64-83 mA, but this improvement was limited at the current further increased. The phosphorus removal in the BEA MBR was attributed to the release of Fe2+, which was closely associated with the applied current to the electrodes. Heme c concentration, enzyme activities, and specific anammox activity exhibited a decreasing trend, while the functional denitrification genes showed a positive correlation with rising voltage. The nitrogen removal efficiency of the BEA system initially increased and then decreased with the voltage rose from 1.5V to 3.5V, peaking at 2.0V of 94.02% ± 1.19%. Transmission electron microscopy and flow cytometry results indicated that accelerated cell apoptosis/lysis led to an irreversible collapse of the biological nitrogen removal system at 3.5V. Candidatus Brocadia was the predominant anammox bacteria in the BEA system. In contrast, closely related Candidatus Kuenenia and Chloroflexi bacteria were gradually eliminated in electrolytic environment. The abundances of Proteobacteria-affiliated denitrifiers were increased with the voltage rising since the organic matter released by the cell apoptosis/lysis was accelerated at a high voltage level.

Innovative approach for predicting biogas production from large-scale anaerobic digester using long-short term memory (LSTM) coupled with genetic algorithm (GA)

An artificial neural network (ANN) model called long-short term memory (LSTM), coupled with a genetic algorithm (GA) for feature selection, was used to predict biogas production of large-scale anaerobic digesters (ADs) of Tehran South Wastewater Treatment Plant (Iran), with a biogas production of approximately 30,000 Nm3/d. In order to employ the real conditions, the hydraulic retention time (HRT) of the ADs (21 days) was considered as the LSTM look-back window. To evaluate the model predictions, three different scenarios were defined. In the first scenario, the model predicted the produced biogas by using raw wastewater characteristics and reached the coefficient of determination of R2 = 0.84. The GA selected four out of eleven parameters of raw wastewater, including loads of BOD5, COD, TSS, and TN (kg/d), as the most informative data for the model. In the second scenario, the model predicted the produced biogas by employing the data of the thickened sludge streams entering the ADs and yielded a higher accuracy (R2 = 0.89). In this scenario, GA selected two out of six parameters of the sludge streams, including total flow rate (m3/d) and average solids content (w/w%). Finally, in the third scenario, by putting the parameters of the two previous scenarios together, the model's prediction accuracy increased slightly (R2 = 0.90). The results demonstrated that the GA-LSTM modeling technique could achieve reliable performance in predicting biogas production of large-scale ADs by including HRT in modeling procedure. It was also found that the raw wastewater characteristics severely affect AD behavior and can be successfully used as the input data of the AD models.

Distinct microbial characteristics of the robust single-stage coupling system during the conversion from anammox-denitritation to anammox-denitratation patterns

Simultaneous anammox-denitrification is effectively operated in two types, i.e., the anammox-denitritation (SAD pattern) and the anammox-denitratation (PDA pattern). The nitrate derived from inevitable nitrite oxidization likely determines the practical operational pattern of the coupling system, while little information is available regarding the microbial characteristics during the pattern conversion. Here, the single-stage bioreactor coupling anammox with denitrification was operated under conditions with a changed ratio of influent nitrite and nitrate. Results showed that the bioreactor exhibited a robust performance during the conversion from SAD to PDA patterns, corresponding with the total nitrogen removal efficiency ranging from 89.5% to 92.4%. Distinct community structures were observed in two patterns, while functional bacteria including the genera Denitratisoma, Thauera, Candidatus Brocadia, and Ca. Jettenia steadily co-existed. Meanwhile, the high transcription of hydrazine synthase genes demonstrated a stable anammox process, while the up-regulated transcription of nitrite and nitrous oxide reductase genes indicated that the complete denitrification process was enhanced for total nitrogen removal during the PDA pattern. Ecologically, stochastic processes dominantly governed the community assembly in two patterns. The PDA pattern improved the interconnectivity of communities, especially for the cooperative behaviors between dominant denitrifying bacteria and low-abundant species. These findings deepen our understanding of the microbial mechanism underlying the different patterns of the coupling system and potentially expand its engineering application.

Machine learning methods for the modelling and optimisation of biogas production from anaerobic digestion: a review

Biogas plant operators often face huge challenges in the monitoring, controlling and optimisation of the anaerobic digestion (AD) process, as it is very sensitive to surrounding changes, which often leads to process failure and adversely affects biogas production. Conventional implemented methods and mechanistic models are impractical and find it difficult to model the nonlinear and intricate interactions of the AD process. Thus, the development of machine learning (ML) algorithms has attracted considerable interest in the areas of process optimization, real-time monitoring, perturbation detection and parameter prediction. This paper provides a comprehensive and up-to-date overview of different machine learning algorithms, including artificial neural network (ANN), fuzzy logic (FL), adaptive network-based fuzzy inference system (ANFIS), support vector machine (SVM), genetic algorithm (GA) and particle swarm optimization (PSO) in terms of working mechanism, structure, advantages and disadvantages, as well as their prediction performances in modelling the biogas production. A few recent case studies of their applications and limitations are also critically reviewed and compared, providing useful information and recommendation in the selection and application of different ML algorithms. This review shows that the prediction efficiency of different ML algorithms is greatly impacted by variations in the reactor configurations, operating conditions, influent characteristics, selection of input parameters and network architectures. It is recommended to incorporate mixed liquor volatile suspended solids (MLVSS) concentration of the anaerobic digester (ranging from 16,500 to 46,700 mg/L) as one of the input parameters to improve the prediction efficiency of ML modelling. This review also shows that the combination of different ML algorithms (i.e. hybrid GA-ANN model) could yield better accuracy with higher R2 (0.9986) than conventional algorithms and could improve the optimization model of AD. Besides, future works could be focused on the incorporation of an integrated digital twin system coupled with ML techniques into the existing Supervisory Control and Data Acquisition (SCADA) system of any biogas plant to detect any operational abnormalities and prevent digester upsets.

Machine learning assisted modelling of anaerobic digestion of waste activated sludge coupled with hydrothermal pre-treatment

This study utilizes decision-tree-based models, including Random Forest, XGBoost, artificial neural networks (ANNs), support vector machine regressors, and K nearest neighbors algorithms, to predict sludge solubilization and methane yield in hydrothermal pretreatment (HTP) coupled with anaerobic digestion (AD) processes. Analyzing two decades of published research, we find that ANN models exhibit superior fitting accuracy for solubilization prediction, while decision-tree models excel in methane yield prediction. Pretreatment temperature is identified as pivotal among various variables, and heating time surprisingly emerges as equally significant as holding time for solubilization and surpasses it for methane yield. Contrary to prior expectations, the HTP method's impact on sludge solubilization and AD performance is minimal. This study underscores data-driven models' potential as resource-efficient tools for optimizing advanced AD processes with HTP. Notably, our research spans nearly two decades of lab, pilot, and full-scale studies, offering novel insights not previously explored.

Microbiome-functionality in anaerobic digesters: A critical review

Microbially driven anaerobic digestion (AD) processes are of immense interest due to their role in the biovalorization of biowastes into renewable energy resources. The function-versatile microbiome, interspecies syntrophic interactions, and trophic-level metabolic pathways are important microbial components of AD. However, the lack of a comprehensive understanding of the process hampers efforts to improve AD efficiency. This study presents a holistic review of research on the microbial and metabolic "black box" of AD processes. Recent research on microbiology, functional traits, and metabolic pathways in AD, as well as the responses of functional microbiota and metabolic capabilities to optimization strategies are reviewed. The diverse ecophysiological traits and cooperation/competition interactions of the functional guilds and the biomanipulation of microbial ecology to generate valuable products other than methane during AD are outlined. The results show that AD communities prioritize cooperation to improve functional redundancy, and the dominance of specific microbes can be explained by thermodynamics, resource allocation models, and metabolic division of labor during cross-feeding. In addition, the multi-omics approaches used to decipher the ecological principles of AD consortia are summarized in detail. Lastly, future microbial research and engineering applications of AD are proposed. This review presents an in-depth understanding of microbiome-functionality mechanisms of AD and provides critical guidance for the directional and efficient bioconversion of biowastes into methane and other valuable products.

End-to-end machine-learning for high-gravity ammonia stripping: Bridging the gap between scientific research and user-friendly applications

The removal and recovery of ammonia from wastewater are critical processes for achieving global environmental sustainability and promoting circular economic development. High-gravity technology is an advanced solution to achieve ammonia stripping from wastewater. This study used machine-learning (ML) techniques to provide more comprehensive insights on various influencing factors, including the operating parameters, wastewater characteristics, and design parameters of rotating packed beds. Bayesian auto-optimization combined with a boosting algorithm effectively overcame the challenges of modeling complex datasets with small sample sizes, multidimensional data, missing values, and skewed distributions. Accurate ML based predictive models for the ammonia removal efficiency (η) and mass transfer coefficient (KLa) were developed, the performance on the training set was R2 = 0.98 and R2 = 0.89, and on the testing set was R2 = 0.98 and R2 = 0.82. The developed model revealed that the stripping stage and gas-liquid ratio were the most influential features for predicting η, whereas the liquid flow and high-gravity factor were the most important features for predicting KLa. The well-trained model was then deployed in an online software application that could provide both predictive and auto-update functions for operators and managers, ensuring that practitioners could use the model. The end-to-end machine-learning approach used in this study-that is, covering data collection, model development, and application-could improve the availability of research results, providing valuable references for the further advancement of technology in the field of environmental.

A Systematic Review of Machine-Learning Solutions in Anaerobic Digestion

The use of machine learning (ML) in anaerobic digestion (AD) is growing in popularity and improves the interpretation of complex system parameters for better operation and optimisation. This systematic literature review aims to explore how ML is currently employed in AD, with particular attention to the challenges of implementation and the benefits of integrating ML techniques. While both lab and industry-scale datasets have been used for model training, challenges arise from varied system designs and the different monitoring equipment used. Traditional machine-learning techniques, predominantly artificial neural networks (ANN), are the most commonly used but face difficulties in scalability and interpretability. Specifically, models trained on lab-scale data often struggle to generalize to full-scale, real-world operations due to the complexity and variability in bacterial communities and system operations. In practical scenarios, machine learning can be employed in real-time operations for predictive modelling, ensuring system stability is maintained, resulting in improved efficiency of both biogas production and waste treatment processes. Through reviewing the ML techniques employed in wider applied domains, potential future research opportunities in addressing these challenges have been identified.

Insights into the recent advances of agro-industrial waste valorization for sustainable biogas production

Recent years have seen a transition to a sustainable circular economy model that uses agro-industrial waste biomass waste to produce energy while reducing trash and greenhouse gas emissions. Biogas production from lignocellulosic biomass (LCB) is an alternative option in the hunt for clean and renewable fuels. Different approaches are employed to transform the LCB to biogas, including pretreatment, anaerobic digestion (AD), and biogas upgradation to biomethane. To maintain process stability and improve AD performance, machine learning (ML) tools are being applied in real-time monitoring, predicting, and optimizing the biogas production process. An environmental life cycle assessment approach for biogas production systems is essential to calculate greenhouse gas emissions. The current review presents a detailed overview of the utilization of agro-waste for sustainable biogas production. Different methods of waste biomass processing and valorization are discussed that contribute towards developing an efficient agro-waste to biogas-based circular economy.

Strategy to enhance the semicontinuous anaerobic digestion of food waste via exogenous additives: experimental and machine learning approaches

The anaerobic digestion (AD) of food waste (FW) was easy to acidify and accumulate ammonia nitrogen. Adding exogenous materials to the AD system can enhance its conversion efficiency by alleviating acidification and ammonia nitrogen inhibition. This work investigated the effects of the addition frequency and additive amount on the AD of FW with increasing organic loading rate (OLR). When the OLR was 3.0 g VS per L per day and the concentration of the additives was 0.5 g per L per day, the stable methane yield reached 263 ± 22 mL per g VS, which was higher than that of the group without the additives (189 mL per g VS). Methanosaetaceae was the dominant archaea, with a maximum abundance of 93.25%. Through machine learning analysis, it was found that the optimal daily methane yield could be achieved. When the OLR was within the range of 0-3.0 g VS per L per day, the pH was within the range of 7.6-8.0, and the additive concentration was more than 0.5 g per L per day. This study proposed a novel additive and determined its usage strategy for regulating the AD of FW through experimental and simulation approaches.

Microbiome data analysis via machine learning models: Exploring vital players to optimize kitchen waste composting system

Composting, reliant on microorganisms, effectively treats kitchen waste. However, it is difficult to precisely understand the specific role of key microorganisms in the composting process by relying solely on experimental research. This study aims to employ machine learning models to explore key microbial genera and to optimize composting systems. After introducing a novel microbiome preprocessing approach, Stacking models were constructed (R2 is about 0.8). The SHAP method (SHapley Additive exPlanations) identified Bacillus, Acinetobacter, Thermobacillus, Pseudomonas, Psychrobacter, and Thermobifida as prominent microbial genera (Shapley values ranging from 3.84 to 1.24). Additionally, microbial agents were prepared to target the identified key genera, and experiments demonstrated that the composting quality score was 76.06 for the treatment and 70.96 for the control. The exogenous agents enhanced decomposition and improved compost quality in later stages. In summary, this study opens up a new avenue to identifying key microorganisms and optimizing the biological treatment process.

Functional conservation of microbial communities determines composition predictability in anaerobic digestion

A major challenge in managing and engineering microbial communities is determining whether and how microbial community responses to environmental alterations can be predicted and explained, especially in microorganism-driven systems. We addressed this challenge by monitoring microbial community responses to the periodic addition of the same feedstock throughout anaerobic digestion, a typical microorganism-driven system where microorganisms degrade and transform the feedstock. The immediate and delayed response consortia were assemblages of microorganisms whose abundances significantly increased on the first or third day after feedstock addition. The immediate response consortia were more predictable than the delayed response consortia and showed a reproducible and predictable order-level composition across multiple feedstock additions. These results stood in both present (16 S rRNA gene) and potentially active (16 S rRNA) microbial communities and in different feedstocks with different biodegradability and were validated by simulation modeling. Despite substantial species variability, the immediate response consortia aligned well with the reproducible CH4 production, which was attributed to the conservation of expressed functions by the response consortia throughout anaerobic digestion, based on metatranscriptomic data analyses. The high species variability might be attributed to intraspecific competition and contribute to biodiversity maintenance and functional redundancy. Our results demonstrate reproducible and predictable microbial community responses and their importance in stabilizing system functions.

A Critical Review of Data Science Applications in Resource Recovery and Carbon Capture from Organic Waste

Municipal and agricultural organic waste can be treated to recover energy, nutrients, and carbon through resource recovery and carbon capture (RRCC) technologies such as anaerobic digestion, struvite precipitation, and pyrolysis. Data science could benefit such technologies by improving their efficiency through data-driven process modeling along with reducing environmental and economic burdens via life cycle assessment (LCA) and techno-economic analysis (TEA), respectively. We critically reviewed 616 peer-reviewed articles on the use of data science in RRCC published during 2002-2022. Although applications of machine learning (ML) methods have drastically increased over time for modeling RRCC technologies, the reviewed studies exhibited significant knowledge gaps at various model development stages. In terms of sustainability, an increasing number of studies included LCA with TEA to quantify both environmental and economic impacts of RRCC. Integration of ML methods with LCA and TEA has the potential to cost-effectively investigate the trade-off between efficiency and sustainability of RRCC, although the literature lacked such integration of techniques. Therefore, we propose an integrated data science framework to inform efficient and sustainable RRCC from organic waste based on the review. Overall, the findings from this review can inform practitioners about the effective utilization of various data science methods for real-world implementation of RRCC technologies.

Hybrid Prediction-Driven High-Throughput Sustainability Screening for Advancing Waste-to-Dimethyl Ether Valorization

Assessing the prospective climate preservation potential of novel, innovative, but immature chemical production techniques is limited by the high number of process synthesis options and the lack of reliable, high-throughput quantitative sustainability pre-screening methods. This study presents the sequential use of data-driven hybrid prediction (ANN-RSM-DOM) to streamline waste-to-dimethyl ether (DME) upcycling using a set of sustainability criteria. Artificial neural networks (ANNs) are developed to generate in silico waste valorization experimental results and ex-ante model the operating space of biorefineries applying the organic fraction of municipal solid waste (OFMSW) and sewage sludge (SS). Aspen Plus process flowsheeting and ANN simulations are postprocessed using the response surface methodology (RSM) and desirability optimization method (DOM) to improve the in-depth mechanistic understanding of environmental systems and identify the most benign configurations. The hybrid prediction highlights the importance of targeted waste selection based on elemental composition and the need to design waste-specific DME synthesis to improve techno-economic and environmental performances. The developed framework reveals plant configurations with concurrent climate benefits (-1.241 and -2.128 kg CO2-eq (kg DME)-1) and low DME production costs (0.382 and 0.492 € (kg DME)-1) using OFMSW and SS feedstocks. Overall, the multi-scale explorative hybrid prediction facilitates early stage process synthesis, assists in the design of block units with nonlinear characteristics, resolves the simultaneous analysis of qualitative and quantitative variables, and enables the high-throughput sustainability screening of low technological readiness level processes.

Intelligent approaches for sustainable management and valorisation of food waste

Food waste (FW) is a severe environmental and social concern that today's civilization is facing. Therefore, it is necessary to have an efficient and sustainable solution for managing FW bioprocessing. Emerging technologies like the Internet of Things (IoT), Artificial Intelligence (AI), and Machine Learning (ML) are critical to achieving this, in which IoT sensors' data is analyzed using AI and ML techniques, enabling real-time decision-making and process optimization. This work describes recent developments in valorizing FW using novel tactics such as the IoT, AI, and ML. It could be concluded that combining IoT, AI, and ML approaches could enhance bioprocess monitoring and management for generating value-added products and chemicals from FW, contributing to improving environmental sustainability and food security. Generally, a comprehensive strategy of applying intelligent techniques in conjunction with government backing can minimize FW and maximize the role of FW in the circular economy toward a more sustainable future.

Machine Learning Approach for Determining and Optimizing Influential Factors of Biogas Production from Lignocellulosic Biomass

Machine learning (ML) was used to predict specific methane yields (SMY) with a dataset of 14 features from lignocellulosic biomass (LB) characteristics and operating conditions of completely mixed reactors under continuous feeding mode. The random forest (RF) model was best suited for predicting SMY with a coefficient of determination (R²) of 0.85 and root mean square error (RMSE) of 0.06. Biomass compositions greatly influenced SMYs from LB, and cellulose prevailed over lignin and biomass ratio as the most important feature. Impact of LB to manure ratio was assessed to optimize biogas production with the RF model. Under typical organic loading rates (OLR), optimum LB to manure ratio of 1:1 was identified. Experimental results confirmed influential factors revealed by the RF model and provided the highest SMY of 79.2% of the predicted value. Successful applications of ML for anaerobic digestion modelling and optimization specifically for LB were revealed in this work.

Prediction of biogas production of industrial scale anaerobic digestion plant by machine learning algorithms

In the anaerobic digestion (AD) process there are some difficulties in maintaining process stability due to the complexity of the system. The variability of the raw material coming to the facility, temperature fluctuations and pH changes as a result of microbial processes cause process instability and require continuous monitoring and control. Increasing continuous monitoring, and internet of things applications within the scope of Industry 4.0 in AD facilities can provide process stability control and early intervention. In this study, five different machine learning (ML) algorithms (RF, ANN, KNN, SVR, and XGBoost) were used to describe and predict the correlation between operational parameters and biogas production quantities collected from a real-scale anaerobic digestion plant. The KNN algorithm had the lowest accuracy in predicting total biogas production over time, while the RF model had the highest prediction accuracy of all prediction models. The RF method produced the best prediction, with an R² of 0.9242, and it was followed by XGBoost, ANN, SVR, and KNN (with R² values of 0.8960, 0.8703, 0.8655, 0.8326, respectively). Real-time process control will be provided and process stability will be maintained by preventing low-efficiency biogas production with the integration of ML applications into AD facilities.

Integrating biokinetics with computational fluid dynamics for energy performance analysis in anaerobic digestion

Anaerobic digestion (AD) is an effective process for decomposing organic matter in wastewater treatment plants (WWTPs) where highly efficient digesters properly mix the sludge. To ensure a uniform substance distribution, a comprehensive modeling method is necessary. Computational fluid dynamics (CFD) helps in the modeling of AD tanks but few studies have focused on integrating hydrodynamics with biokinetics because of complex AD processes. The current study presents a new CFD platform for estimating the biokinetics of WWTPs to assess the energy performance of AD tanks. The presented method is validated by numerical and experimental studies, and facilitates a link between methane production and mixing energy consumption. The on-site settings of the recirculation mixing system in the studied WWTP was able to prepare a uniform mixture of the material. However, reducing mixing rate to decrease energy consumption did not lead to proper mixing quality.

Enhanced anaerobic digestion performance of food waste by zero-valent iron and iron oxides nanoparticles: Comparative analyses of microbial community and metabolism

The effects of zero-valent iron (ZVI) and iron oxides nanoparticles on anaerobic digestion (AD) performance of food waste (FW) were comparably clarified in this study. Results indicated that the nanoparticles supplement effectively enhanced the methane yields. As observed, these nanoparticles accelerated organics transformation and alleviated acidification process. Also, the enriched total methanogens and functional bacteria (e.g., Proteiniphilum) were consistent with the promotion of oxidative phosphorylation, citrate cycle, coenzymes biosynthesis and the metabolisms of amino acid, carbohydrate, methane. Additionally, these nanoparticles stimulated electron transfer potential via enriching syntrophic genera (e.g., Geobacter, Syntrophomonas), primary acetate-dependent methanogens (Methanosaeta, Methanosarcina) and related functions (pilus assembly protein, ferredoxins). By comparison, ZVI nanoparticle presented the excellent performance on methanogenesis. This study provides comprehensive understanding of the methanogenesis facilitated by ZVI and iron oxides nanoparticles through the enhancement of key microbes and microbial metabolisms, while ZVI is an excellent option for promoting the methane production.

Development of an artificial intelligence-based framework for biogas generation from a micro anaerobic digestion plant

Despite the advantages of the Anaerobic Digestion (AD) technology for organic waste management, low system performance in biogas production negatively affects the wide spread of this technology. This paper develops a new artificial intelligence-based framework to predict and optimise the biogas generated from a micro-AD plant. The framework comprises some main steps including data collection and imputation, recurrent neural network/ Non-Linear Autoregressive Exogenous (NARX) model, shuffled frog leaping algorithm (SFLA) optimisation model and sensitivity analysis. The suggested framework was demonstrated by its application on a real micro-AD plant in London. The NARX model was developed for predicting yielded biogas based on the feeding data over preceding days in which their lag times were fine-tuned using the SFLA. The optimal daily feeding pattern to obtain maximum biogas generation was determined using the SFLA. The results show that the developed framework can improve the productivity of biogas in optimal operation strategy by 43 % compared to business as usual and the average biogas produced can raise from 3.26 to 4.34 m³/day. The optimal feeding pattern during a four-day cycle is to feed over the last two days and thereby reducing the operational costs related to the labour for feeding the plant in the first two days. The results of the sensitivity analysis show the optimised biogas generation is strongly influenced by the content of oats and catering waste as well as the optimal allocated day for adding feed to the main digester compared to other feed variables e.g., added water and soaked liner.

Machine learning and circular bioeconomy: Building new resource efficiency from diverse waste streams

Biorefinery systems are playing pivotal roles in the technological support of resource efficiency for circular bioeconomy. Meanwhile, artificial intelligence presents great potential in handling scientific tasks of high-dimensional complexity. This review article scrutinizes the status of machine learning (ML) applications in four critical biorefinery systems (i.e. composting, fermentation, anaerobic digestion, and thermochemical conversions) as well as their advancements against traditional modeling techniques of mechanistic approach. The contents cover their algorithm selections, modeling challenges, and prospective improvements. Perspectives are sketched to further inform collective efforts on crucial aspects. The multidisciplinary interchange of modeling knowledge will enable a more progressive digital transformation of sustainability efforts in supporting sustainable development goals.

Applications of artificial intelligence in anaerobic co-digestion: Recent advances and prospects

Anaerobic co-digestion (AcoD) offers several merits such as better digestibility and process stability while enhancing methane yield due to synergistic effects. Operation of an efficient AcoD system, however, requires full comprehension of important operational parameters, such as co-substrates ratio, their composition, volatile fatty acids/alkalinity ratio, organic loading rate, and solids/hydraulic retention time. AcoD process optimization, prediction and control, and early detection of system instability are often difficult to achieve through tedious manual monitoring processes. Recently, artificial intelligence (AI) has emerged as an innovative approach to computational modeling and optimization of the AcoD process. This review discusses AI applications in AcoD process optimization, control, prediction of unknown input/output parameters, and real-time monitoring. Furthermore, the review also compares standalone and hybrid AI algorithms as applied to AcoD. The review highlights future research directions for data preprocessing, model interpretation and validation, and grey-box modeling in AcoD process.

Review of explainable machine learning for anaerobic digestion

Anaerobic digestion (AD) is a promising technology for recovering value-added resources from organic waste, thus achieving sustainable waste management. The performance of AD is dictated by a variety of factors including system design and operating conditions. This necessitates developing suitable modelling and optimization tools to quantify its off-design performance, where the application of machine learning (ML) and soft computing approaches have received increasing attention. Here, we succinctly reviewed the latest progress in black-box ML approaches for AD modelling with a thrust on global and local model interpretability metrics (e.g., Shapley values, partial dependence analysis, permutation feature importance). Categorical applications of the ML and soft computing approaches such as what-if scenario analysis, fault detection in AD systems, long-term operation prediction, and integration of ML with life cycle assessment are discussed. Finally, the research gaps and scopes for future work are summarized.

Advances in machine learning for high value-added applications of lignocellulosic biomass

Lignocellulose can be converted into biofuel or functional materials to achieve high value-added utilization. Biomass utilization process is complex and multi-dimensional. This paper focuses on the biomass conversion reaction conditions, the preparation of biomass-based functional materials, the combination of biomass conversion and traditional wet chemistry, molecular simulation and process simulation. This paper analyzes the mechanism, advantages and disadvantages of important machine learning (ML) methods. The application examples of ML in different aspects of high value utilization of lignocellulose are summarized in detail. The challenges and future prospects of ML in this field are analyzed.

Machine learning in fermentative biohydrogen production: Advantages, challenges, and applications

Hydrogen can be produced in an environmentally friendly manner through biological processes using a variety of organic waste and biomass as feedstock. However, the complexity of biological processes limits their predictability and reliability, which hinders the scale-up and dissemination. This article reviews contemporary research and perspectives on the application of machine learning in biohydrogen production technology. Several machine learning algorithems have recently been implemented for modeling the nonlinear and complex relationships among operational and performance parameters in biohydrogen production as well as predicting the process performance and microbial population dynamics. Reinforced machine learning methods exhibited precise state prediction and retrieved the underlying kinetics effectively. Machine-learning based prediction was also improved by using microbial sequencing data as input parameters. Further research on machine learning could be instrumental in designing a process control tool to maintain reliable hydrogen production performance and identify connection between the process performance and the microbial population.

Prediction and optimization of medium-chain carboxylic acids production from food waste using machine learning models

Machine learning models were developed in this study to predict and optimize the medium-chain carbolic acids (MCCAs) production from food waste. All three selected prediction algorithms achieved decent performance (accuracy > 0.85, R² > 0.707). Three optimization algorithms were applied for MCCA production optimization based on the prediction algorithms. The maximum MCCA production rate (0.68 g chemical oxygen demand per liter per day) was achieved by simulated annealing coupled with random forest under the optimal conditions of pH 8.3, temperature 50 °C, retention time 4 days, loading rate 15.8 g volatile solid per liter per day, and inoculum to food waste ratio 70:30 with semi-continuous mode. Further experiments validated (18 % error) that the MCCA production rate was 113 % higher than the highest production rate of current lab experiments and 60 % higher than the statistical optimization using response surface methodology. This study demonstrates the potential of using machine learning for MCCA production prediction and optimization.

Integration of anaerobic digestion with heat Pump: Machine learning-based technical and environmental assessment

Anaerobic digestion (AD)-based biogas production mitigates the environmental footprint of organic wastes (e.g., food waste and sewage sludge) and facilitates a circular economy. The work proposed an integrated system where the thermal energy demand of an AD is supplied using an air source heat pump (ASHP). The proposed system is compared to a baseline system, where the thermal energy is supplied by a natural gas-based heating system. Several machine learning models are developed for predicting biogas production, among which the Gaussian Process Regression (GPR) showed a superior performance (R² = 0.84 and RMSE = 0.0755 L gVS^-1 day^-1). The GPR model further informed a thermodynamic model of the ASHP, which revealed the maximum biogas yield to be approximately 0.585 L.gVS^-1.day^-1 at an optimal temperature of 55 °C (thermophilic). Subsequently, life cycle assessment showed that ASHP-based AD heating systems achieved 28.1 % (thermophilic) and 36.8 % (mesophilic) carbon abatement than the baseline system.

Plant-scale biogas production prediction based on multiple hybrid machine learning technique

The parameters from full-scale biogas plants are highly nonlinear and imbalanced, resulting in low prediction accuracy when using traditional machine learning algorithms. In this study, a hybrid extreme learning machine (ELM) model was proposed to improve prediction accuracy by solving imbalanced data. The results showed that the best ELM model had a good prediction for validation data (R² = 0.972), and the model was developed into the software (prediction error of 2.15 %). Furthermore, two parameters within a certain range (feed volume (FV) = 23-45 m³ and total volatile fatty acids of anaerobic digestion (TVFA_AD) = 1750-3000 mg/L) were identified as the most important characteristics that positively affected biogas production. This study combines machine learning with data-balancing techniques and optimization algorithms to achieve accurate predictions of plant biogas production at various loads.

Model development of bioelectrochemical systems: A critical review from the perspective of physiochemical principles and mathematical methods

Bioelectrochemical systems (BESs) are promising devices for wastewater treatment and bio-energy production. Since various processes are interacted and affect the overall performance of the device, the development of theoretical modeling is an efficient approach to understand the fundamental mechanisms that govern the performance of the BES. This review aims to summarize the physiochemical principle and mathematical method in BES models, which is of great importance for the establishment of an accurate model while has received little attention in previous reviews. In this review, we begin with a classification of existing models including bioelectrochemical models, electronic models, and machine learning models. Subsequently, physiochemical principles and mathematical methods in models are discussed from two aspects: one is the description of methodology how to build a framework for models, and the other is to further review additional methods that can enrich model functions. Finally, the advantages/disadvantages, extended applications, and perspectives of models are discussed. It is expected that this review can provide a viewpoint from methodologies to understand BES models.