A5019533763- Environmental Impact and Sustainability
- Green IT and Sustainability
- Electric Vehicles and Infrastructure
- Energy, Environment, and Transportation Policies
- Smart Grid Energy Management
- Recycling and Waste Management Techniques
- Climate Change Policy and Economics
- Building Energy and Comfort Optimization
- Sustainable Industrial Ecology
- Cloud Computing and Resource Management
- Software-Defined Networks and 5G
- Advanced Data Storage Technologies
- Biofuel production and bioconversion
- Distributed and Parallel Computing Systems
- Explainable Artificial Intelligence (XAI)
- Energy Load and Power Forecasting
- Global Energy and Sustainability Research
- Caching and Content Delivery
- Energy and Environment Impacts
- Renewable Energy and Sustainability
- Social Acceptance of Renewable Energy
- Transportation and Mobility Innovations
- Photovoltaic Systems and Sustainability
- Sustainable Building Design and Assessment
- Power Line Communications and Noise
Polytechnique Montréal
2011-2020
Université de Montréal
2014-2018
École de Technologie Supérieure
2014-2016
Ericsson (Canada)
2014
From an environmental standpoint, there are a few crucial aspects of training neural network that have major impact on the quantity carbon it emits. These factors include: location server used for and energy grid uses, length procedure, even make model hardware which takes place. In order to approximate these emissions, we present our Machine Learning Emissions Calculator, tool community better understand ML models. We accompany this with explanation cited above, as well concrete actions...
Summary The divide between attributional and consequential research perspectives partly overlaps with the long‐standing methodological discussions in life cycle assessment (LCA) input‐output analysis (IO) communities on choice of techniques models for dealing situations coproduction. recent harmonization LCA allocations IO constructs revealed a more diverse set coproduction than had previously been understood. This increased flexibility transparency inventory modeling warrants re‐evaluation...
Summary The environmental impacts of data centers that provide information and communication technologies (ICTs) services are strongly related to electricity generation. With the increasing use ICT, many expected be built, causing more absolute on environment. Given distribution networks very complex dynamic systems, an evaluation future is uncertain. This study proposes a new approach investigate consequences center deployment in Canada optimize this based Energy 2020 technoeconomic model...
Smart homes play a crucial role in reducing the residential sector electricity consumption and Greenhouse Gases (GHG) emissions. In this work, we present time series approach to predict GHG emissions be integrated into smart home management systems. More specifically, used Long Short-Term Memory (LSTM), variant of Recurrent Neural Networks. The prediction results get mean absolute percentage error (MAPE) close 2 percent when region under study has an energy matrix mostly based on fossil...
The information and communications technologies (ICT) sector is seeking to reduce the electricity consumption of data processing centres.Among initiatives improve energy efficiency shift cloud computing technology.Thanks very favourable geographical conditions, Canadian mix highly suited implementation centres, especially in light significant potential renewable energy, which can help curb greenhouse gas emissions.In green sustainable Telco (GSTC) project, an efficient network would be set...
Information and communication technologies are believed to contribute reduce environmental impacts in other economic sectors. Smart homes aim this purpose through improved resource management, but actual benefits uncertain. Indeed, an increase efficiency may lead consumption thus impacts, a phenomenon called rebound effect. Here we describe agent-based approach which simulates human behaviors during the use phase of smart allows computation including those due In our simulation, impact by...
Summary Many countries see biofuels as a replacement to fossil fuels mitigate climate change. Nevertheless, some concerns remain about the overall benefits of policies. More comprehensive tools seem required evaluate indirect effects biofuel This article proposes method large‐scale policies that is based on life cycle assessment (LCA), environmental extensions input‐output (I‐O) tables, and general equilibrium model. The enables policies, including land‐use changes (LUCs), in context...
In this work, three of many ICT-specific challenges LCA are discussed.First, the inconsistency versus uncertainty is reviewed with regard to meta-technological nature ICT.As an example, semiconductor technologies used highlight complexities especially respect energy and water consumption.The need for specific representations metric separately assess products discussed.It highlighted that applying product-oriented approaches would result in abandoning or disfavoring new could otherwise help...
Recently, data centres have been called out for their particularly high energy consumption, which already accounts 1.5% of the total global electricity consumption and is among world's fastest growing consumptions.To reduce centres' environmental impacts, technologies such as free cooling sustainable power sources are used.Another newly developed strategy to improve efficiency virtualization, makes it possible install several operating systems, known virtual machines (VMs), so that tasks...
Among the innovative approaches to reduce greenhouse gas (GHG) emissions of data centres during their use phase, electrical power from renewable sources appears promising. However, electricity is often intermittent due meteorological conditions. Consequently, regional availability varies constantly over time. This created opportunity deploy cloud computing systems relying on located in different regions. Cloud technology enables real-time load migration a centre region where GHG per kWh are...