An AI Climate Conversation

Data centres enable internet and cloud services and contribute over 3% of global greenhouse gas (GHG) emissions. This is more than the share of the aviation industry. With rising demand driven by AI and digitization, this share is projected to grow exponentially. Innovative solutions to curb emissions in this sector are critical to address our planet’s climate crisis. A key obstacle for innovation is a lack of high‐quality end‐to‐end GHG emissions data. Today, it is impossible to validate the carbon figures reported by cloud providers. This is because all actors in this ecosystem have incentives to under‐report emissions to appear sustainable. 

Building an end‐to‐end tracking system for storing emissions data is crucial, as modern distributed applications often use numerous services and operate across local infrastructure and/or one or more cloud domains. Lack of end‐to‐end data hampers GHG attribution to specific cloud users/apps. Resolving this drives measurable action from stakeholders: providers, developers, and end‐users. 

This project addresses this significant challenge by developing a tamper‐resistant decentralized GHG emission tracking framework. Our objective is to design custom blockchain‐based protocols and tracking methods that safeguard proprietary information of vendors while maintaining the integrity of emissions reporting. The decentralized nature of this system mitigates the potential for collusion among a minority of participants. Furthermore, this system will facilitate interoperability, enabling carbon tracking even when an application’s execution transitions between on‐premise and cloud environments or crosses multiple cloud platforms. 

Project Mentors

Ifeoma Adaji, Computer Science, Irving K. Barber Faculty of Science 
Niyi Asiyanbi, Geography, Irving K. Barber Faculty of Arts and Social Sciences 

Solution Scholars

Alfie Brazier, Institute for Resources, Environment and Sustainability, Faculty of Science
Emon Sarker, Computer Science, Irving K. Barber Faculty of Science

Project Description 

 Wildfires are an increasing threat due to climate change, and their frequency and intensity pose significant risks to communities, infrastructure, and ecosystems. Despite the availability of wildfire preparedness guidelines, many individuals remain unprepared due to lack of awareness, accessibility issues, and psychological barriers to proactive planning. This project will build on the foundation of the WISEC app which was designed to support wildfire preparedness in the city of Whistler, British Columbia using feedback from frontline fire management stakeholders. The proposed project aims to enhance WISEC's impact by integrating new features that address behavioral engagement, accessibility, and community resilience for all cities in BC. Additionally, we will develop a web‐based version of the application to expand accessibility and usability across devices. The proposed new version of the mobile app and the web application will be available to all cities in BC.  The project seeks to accomplish three goals:

1. Expand accessibility by developing a web‐based version of the WISEC app to complement the mobile app, ensuring broader usability for different demographics, including those without smartphones.
2. Integrate AI‐driven insights to the current mobile application to provide real‐time risk assessments, emergency alerts, and tailored recommendations based on user location and readiness level.
3. Extend the use case of the app from Whistler to the rest of BC.