Analysis of input set characteristics and variances on k-fold cross validation for a Recurrent Neural Network model on waste disposal rate estimation.

The use of machine learning techniques in waste management studies is increasingly popular. Recent literature suggests k-fold cross validation may reduce input dataset partition uncertainties and minimize overfitting issues. The objectives are to quantify the benefits of k-fold cross validation for municipal waste disposal prediction and to identify the relationship of testing dataset variance on predictive neural network model performance. It is hypothesized that the dataset characteristics and variances may dictate the necessity of k-fold cross validation on neural network waste model construction. Seven RNN-LSTM predictive models were developed using historical landfill waste records and climatic and socio-economic data. The performance of all trials was acceptable in the training and validation stages, with MAPE all less than 10%. In this study, the 7-fold cross validation reduced the bias in selection of testing sets as it helps to reduce MAPE by up to 44.57%, MSE by up to 54.15%, and increased R value by up to 8.33%. Correlation analysis suggests that fewer outliers and less variance of the testing dataset correlated well with lower modeling error. The length of the continuous high waste season and length of total high waste period appear not important to the model performance. The result suggests that k-fold cross validation should be applied to testing datasets with higher variances. The use of MSE as an evaluation index is recommended.

The use of a recurrent neural network model with separated time-series and lagged daily inputs for waste disposal rates modeling during COVID-19.

A new modeling framework is proposed to estimate mixed waste disposal rates in a Canadian capital city during the pandemic. Different Recurrent Neural Network models were developed using climatic, socioeconomic, and COVID-19 related daily variables with different input lag times and study periods. It is hypothesized that the use of distinct time series and lagged inputs may improve modeling accuracy. Considering the entire 7.5-year period from Jan 2013 to Sept 2020, multi-variate weekday models were sensitive with lag times in the testing stage. It appears that the selection of input variables is more important than waste model complexity. Models applying COVID-19 related inputs generally had better performance, with average MAPE of 10.1%. The optimized lag times are however similar between the periods, with slightly longer average lag for the COVID-19 at 5.3 days. Simpler models with least input variables appear to better simulate waste disposal rates, and both 'Temp-Hum' (Temperature-Humidity) and 'Temp-New Test' (Temperature-COVID new test case) models capture the general disposal trend well, with MAPE of 10.3% and 9.4%, respectively. The benefits of the use of separated time series inputs are more apparent during the COVID-19 period, with noticeable decrease in modeling error.

Modeling of municipal waste disposal rates during COVID-19 using separated waste fraction models.

Municipal waste disposal behaviors in Regina, the capital city of Saskatchewan, Canada have significantly changed during the COVID-19 pandemic. About 7.5 year of waste disposal data at the Regina landfill was collected, verified, and consolidated. Four modeling approaches were examined to predict total waste disposal at the Regina landfill during the COVID-19 period, including (i) continuous total (Baseline), (ii) continuous fraction, (iii) truncated total, and (iv) truncated fraction. A single feature input recurrent neural network model was adopted for each approach. It is hypothesized that waste quantity modeling using different waste fractions and separate time series can better capture disposal behaviors of residents during the lockdown. Compared to the baseline approach, the use of waste fractions in modeling improves both result accuracy and precision. In general, the use of continuous time series over-predicted total waste disposal, especially when actual disposal rates were less than 50 t/day. Compared to the baseline approach, mean absolute error (MAE), mean absolute percentage error (MAPE), and mean square error (MSE) were reduced. The R value increased from 0.63 to 0.79. Comparing to the baseline, the truncated total and the truncated fraction approaches better captured the total waste disposal behaviors during the COVID-19 period, probably due to the periodicity of the weeklong data set. For both approaches, MAE and MAPE were lower than 70 and 22%, respectively. The model performance of the truncated fraction appears the best, with an MAPE of 19.8% and R value of 0.92. Results suggest the uses of waste fractions and separated time series are beneficial, especially if the input set is heavily skewed.

Identification of behaviour patterns in waste collection and disposal during the first wave of COVID-19 in Regina, Saskatchewan, Canada.

The novel coronavirus (2019-nCov) has had significant impacts on almost every aspect of daily life. From 'stay-at-home' orders to the progressive lifting of restrictions, the COVID-19 pandemic has had unprecedented effects on consumer behaviours and waste disposal habits. The purpose of this short communication is to examine time series waste collection and disposal data in a mid-sized Canadian city to understand how behavioural changes have affected municipal waste management. The results suggest that private waste disposal increased during the pandemic. This may be due to people doing home renovations in order to accommodate working from home. Furthermore, it appears that changes in consumer habits destabilized the consistency of waste disposal tonnage when compared to the same time period in 2019. When considering curbside residential waste collection, there was also an increase in tonnage. This may be the result of more waste being generated at home due to changes in eating and cooking habits, and cleaning routine. Finally, the ratio of residential waste collection to total disposal is examined. More residential waste is being generated, which may have environmental and operational effects, especially related to collection and transportation. The results from this study are important from an operational perspective, and will help planners and policy makers to better prepare for changes in the waste stream due to pandemics or other emergencies.

Waste disposal characteristics and data variability in a mid-sized Canadian city during COVID-19.

COVID-19, declared a global pandemic by the World Health Organization, has caused governments to react swiftly with a variety of measures to quell the spread of the virus. This study investigates changes in waste disposal characteristics and the relationship between the mass of biomedical waste disposed and new COVID-19 tests performed in Regina, Canada. Results suggest that between May and September 2020, significant differences in the median amount of waste disposed exist. The amount of monthly waste disposed was slightly lower to about 450-550 tonnes/month. Monthly waste data variability, however, was significantly lower. Seasonal effects on total waste disposal is observed, but is less obvious than pre-COVID time. Furthermore, the distribution of different waste fractions varies, probably due to operational and industrial characteristics. A non-linear relationship exists between the number of COVID-19 tests performed and the mass of biomedical waste disposed, perhaps due to a lagged relationship between biomedical waste generation and disposal.

Change in Compliance of Staff at 4 Vietnam Universities after the Enactment of Smoke-free Environment Decree.

BACKGROUND: The Tobacco Control Law in Vietnam was adopted in 2012 and came into force from May 2013, followed by a number of guiding sub-law legal documents. Smoke-free campus policy in university is considered an important measure to protect people from secondhand smoking as well as staffs and students will be in favour of the policy. Furthermore, there has been evidence suggested that smoke-free policy had positive impact on active smoking as well as anti-smoking attitude. METHODS: A cross-sectional self-administered study of staff at 4 universities were conducted in 2 phases, Phase 1 as early introduction of the Law (n = 900) and Phase 2 as 1-year post (n = 885). Demographics, tobacco consumption, compliance status and compliance with awareness towards the campus smoking regulations were assessed in both phases. RESULTS: Daily smoking prevalence decreased significantly (P < .05) 1 year after implementing the smoke-free policy, while a significant increase in occasional smoking (P < .01). Compliance of staff to the regulation the campus should be indoor smoke-free was significantly increase at Phase 2 compared to Phase 1, however participants reported there would be places on campus that smokers frequently violated the smoke-free regulations. CONCLUSIONS: The study indicated a significant positive change in compliance of staff at the 4 universities after the implementation of the Tobacco Control Law, included the smoke-free policy. Although the prevalence of tobacco smoking in this study was low, the proportion of respondents who reported to reduce infringement the smoke-free policy suggests support for staff smokers would be beneficial. Raising awareness and enforcement is likely to enhance the long-term outcomes of the implementation of smoke-free environment.

Application of a multi-stage neural network approach for time-series landfill gas modeling with missing data imputation.

To mitigate the greenhouse gas effect, accurate and precise landfill gas prediction models are required for more precise prediction of the amount and recovery time of methane gas from landfills. When the study associates to greenhouse gas emissions problems, time series prediction models are of considerable interests, in which significant past records of gas data are required. This study is the first to specially impute the missing methane (CH4) data for applying in time series artificial neural network (ANN) model in an attempt to predict daily CH4 generation rate from a landfill in Regina, SK, Canada. Pre-processing was conducted on data to evaluate independent and significant meteorological input variables and provide suitable dataset for developing CH4 generation models. A two-stage time series model proposed in this study was performed by missing data imputation at the first stage, followed by a neural network auto-regressive model with exogenous inputs (NARX) at the second stage. The model with 3 layers, 5 climatic variables and 9 neurons in the hidden layer was the optimal structure. This model shows the high performance in CH4 prediction with the average index of agreement of 0.92 and the average mean absolute percentage error (MAPE) of 3.03% during the testing stage. Missing data imputation coupled with NARX method decreased the mean squared error (MSE) of the model by 84% (compared to Multilayer Perceptrons neural network model) in the testing period representing the effectiveness of missing data estimation coupling with time series ANN models in daily CH4 generation prediction.

Interactions of residential waste composition and collection truck compartment design on GIS route optimization.

Waste collection is an important functional element in a modern waste management system; and may account for up to half of the total expenditure on waste management in industrialized nations. Most optimization of waste collection studies include truck route distance and fuel consumption considerations without explicitly considering the inter-relationships of the model parameters. This study however delineates the complex inter-relationships of waste composition, collection frequency, collection type, and truck compartment configurations in a small waste collection zone in Austin, Texas. A total of 48 different scenarios are modelled and investigated. Truck travel distances are found sensitive to collection frequency, truck capacity, volume ratio of truck compartment, and waste density. The results showed that the increase in waste density and waste collection frequency helped to save up to 18.2% in travel distances and 41.9% in travel time. Waste composition is significant in travel distance, regardless of truck design. Increasing truck capacity by 25% helped to save 4.1-24.4% of truck travel distances. Optimal volume ratio of truck compartments was 50:50 (50% volume for garbage and 50% volume for recyclables); a finding that is different than what is currently reported in the literature; pointing to the site-specific nature of studies of this type. The use of dual compartment trucks helps to reduce travel distances by up to 23.0% and travel time by up to 14.3%. It appears that the minimization of operation time within the collection area is key to an efficient system.

Assessment of waste characteristics and their impact on GIS vehicle collection route optimization using ANN waste forecasts.

Combining an artificial neural network (ANN) waste prediction model with a geographic information system (GIS) waste collection route optimization, the paper shows how the compositional features of waste materials affect the optimized truck route time, distance, and air emissions. Using data from Austin, Texas, USA, a nonlinear autoregressive ANN model is used to predict the waste generation rate of the recycling and garbage streams for the year 2023 in four sub-areas of the city. This ANN model resulted in mean absolute percentage errors ranging from 10.92% to 16.51%. Modified compositions of the recycling and garbage streams are then used as inputs, along with the year 2023 generation rates, to create 6 modified and 3 non-modified scenarios that reflect possible future changes in waste composition. These waste stream scenarios are then used as input parameters to determine optimal waste collection routes with minimal travel distance in each of the four sub-areas using the GIS vehicle routing problem network analysis tool. Results of these 36 scenarios yield changes in travel distance of up to 19.9%, when compared to the non-modified composition. Further, dual compartment trucks were compared to single compartment trucks and found to save between 10.3 and 16.0% in travel distance and slightly reduce emissions but had a 15.7-19.8% increase in collection time. Results suggest temporal changes in waste composition and characteristics are important in GIS route optimization studies.

Time-lagged effects of weekly climatic and socio-economic factors on ANN municipal yard waste prediction models.

Efficient and effective solid waste management requires sufficient ability to predict the operational capacity of a system correctly. Waste prediction models have been widely studied and these models are always being challenged to perform more accurately. Unlike waste prediction models for mixed wastes, variables for yard waste are time sensitive and the effects of lag must be explicitly considered. This study is the first to specifically look at lag times relating to variables that attempt to predict municipal yard waste generation using machine learning approaches. Weekly averaged climatic and socio-economic variables are screened through correlation analysis and the significant variables are then used to develop yard waste models. These models then utilize artificial neural networks (ANN) where the variables are time lagged for a different number of weeks. This helps to realize a reduction in the error of the predicted weekly yard waste generation. Optimal lag times for each model varied from 1 to 11 weeks. The best model used both the ambient air temperature and population variables, in an ANN model with 3 layers, 11 neurons in the hidden layer, and an optimal lag time of 1 week. A mean absolute percentage error of 18.72% was obtained during the testing stage. One model saw a 55.4% decrease in the mean squared error at training, showing the value of lag time on the accuracy of weekly yard waste prediction models.

Use of seasonal parameters and their effects on FOD landfill gas modeling.

Temporal and spatial variations in landfill gas generations and emissions have been observed and reported by others. Real-time gas data between 2008 and 2014 from a municipal landfill located in a cold, semi-arid climate were consolidated to fit a linear-interpolated form of LandGEM. Seasonal variations in gas collection were observed in the landfill. LandGEM's default decay rate k was not applicable for this Canadian landfill due to significant overestimation (32.2% error). Optimal seasonal k and Lo collection parameters had 8.1% error compared to field data, compared to 8.3% error using optimal annual parameters. The optimal kwinter was 0.0118 year-1 and the ksummer was 0.0141 year-1 (14.7% difference), with a corresponding Lo of 100.0 m3/Mg which changed negligibly between the sets. Three pseudo-second order iterative methods were considered, and evaluated using RSS and generation parameters in the literature. A simple application study was conducted using LFGcost-Web, and found the increased precision of seasonal k's resulted in negligible differences with annual optimized k. The default parameters overestimated the net present worth by 12-155% for three of the four common LFG energy projects.

Parameter interrelationships in a dual phase GIS-based municipal solid waste collection model.

Geographic information systems are a valuable tool for waste collection and optimization, but they have been underutilized in helping to understand the complex interrelationships that exist within a dual phase solid waste collection system. A GIS-based dual phase model integrating the handcart pre-collection phase and truck collection phase for a study area located in Hai Phong, Vietnam was proposed, and a resulting total system cost was estimated. Temporary collection points were first identified using both the maximize coverage and minimize facility location-allocation tools from a list of candidate temporary collection points and constraints. Two vehicle routing problems were then separately modeled for handcart and truck routes. A total of 30 scenarios were considered in order to investigate the interrelationships between the model parameters, with respect to the total operation costs and maintenance system costs. The scenario with 11 temporary collection points and a maximum handcart collection distance of 500 m gave the lowest overall cost in the study area. The results suggest a single temporary collection point in the study is able to serve about 2,590 people in an area of 0.11 km2. Compared to the status quo condition, a 13.76% reduction in truck travel distances is attainable using the proposed model. It is found that the number and distribution of temporary collection points greatly affected the cost effectiveness in both pre-collection and collection phases.

Optimization of first order decay gas generation model parameters for landfills located in cold semi-arid climates.

Canada has one of the highest waste generation rates in the world. Because of high land availability, land disposal rates in the province of Saskatchewan are high compared to the rest of the country. In this study, landfill gas data was collected at semi-arid landfills in Regina and Saskatoon, Saskatchewan, and curve fitting was carried out to find optimal k and Lo or DOC values using LandGEM, Afvalzorg Simple, and IPCC first order decay models. Model parameters at each landfill were estimated and compared using default k and Lo or DOC values. Methane generation rates were substantially overestimated using default values (with percentage errors from 55 to 135%). The mean percentage errors for the optimized k and Lo or DOC values ranged from 11.60% to 19.93% at the Regina landfill, and 1.65% to 10.83% at the Saskatoon landfill. Finally, the effect of different iterative methods on the curve fitting process was examined. The residual sum of squares for each model and iterative approaches were similar, with the exception of iterative method 1 for the IPCC model. The default values in these models fail to represent landfills located in cold semi-arid climates. The use of site specific data, provided enough information is available regarding waste mass and composition, can greatly help to improve the accuracy of these first order decay models.