Back in 2015, the United Nation Climate Change Conference led to the Paris Agreement setting the goal of limiting global warming to “well below 2 °C” compared to pre-industrial levels was signed. To support this objective, in 2021, the Science-Based Target Initiative (SBTI) set a demanding scientific objective to keep global warming at 1.5°C. Hence reducing our carbon footprint has become a core challenge that not only individuals but corporations have to engage with in all their operations. And R&D has a pivotal role in making it happen.
Why Assess for Carbon Neutrality?
Therefore, most companies have incorporated the global warming stake as one of the main pillars of their corporate social responsibility. To do so, for example, the Atos Group’s has set up its Environmental Management System (EMS) – ISO 14001 certified- has been aligned to enable its customers and employees to live, work and develop sustainably, in a safe and secure information space.
All elements linked with the activities, products or services of an organization that can impact the environment should be assessed to conclude the mitigation steps to be initiated, if any.
An example of impacts pertinent to our domain is consumption of energy that comes from fossil fuels thereby leading to carbon emissions.
So, it is imperative that assessment is done to improve the environmental performance of digital solutions and technologies in the fight against global warming.
The case of Cloud MDR operations
The MDR service that Eviden provides worldwide has been certified to be carbon neutral. In this article we aim to explain the main dimensions used for assessing the carbon neutral status and the process that was followed for computing it.
The Managed Detection and Response (MDR) Service consists of two parts – the people who provide the SOC services and the technology represented by the AIsaac platform that is used to deliver the service. Both these parts were considered to arrive at the total carbon emissions for the MDR service.
Eviden’s MDR service benefits from an eco-design that leverages a cloud environment for the platform and operational efficiency achieved through usage of the AIsaac platform. AIsaac also uses AI & Automation to improve operational efficiencies and thus helps in reducing the number of people required to run the MDR service.
How to Assess for Carbon Neutrality?
An assessment of Carbon Neutrality starts with identifying a service / product of the company that may cause environmental pollution either directly or indirectly. Then the specific elements of the service / product are identified that could have an environmental impact. For example, in the case of the MDR Service we identified two elements – manpower and technology in the form of the AIsaac platform that is used to deliver the service.
This is followed by an assessment of the carbon emissions produced by the elements identified. Based on the type of energy consumption of the technology used in the service/product, a CO2 footprint is calculated. If the service / product is found to be producing residual non-zero carbon emissions, then suitable mitigation action in the form of CO2 credits that will be required is arrived at. To make the service / product carbon neutral, there will be a need to contribute money (to the extent of the CO2 credits required) to reputed organizations that are at the forefront of this global carbon neutrality program. Such monetary contributions help these organizations achieve a lower carbon emission in other parts of the earth. Thus, looked in its totality, the service / product becomes carbon neutral.
The Process for MDR Service Assessment
The exercise of assessing the Carbon Neutrality of the MDR service involved the following stages:
Scoping & Strategy
Assessment & Calculations
Purchase of Carbon Credits required
1. Scoping & Strategy
The Cloud based MDR Service was identified as the scope for the carbon neutral assessment. Also the two main components for delivering the services – People and the AIsaac platform – were identified as the driving components of the service which needed assessment.
We created multiple approaches for gathering data on cloud resources used in the service delivery and the number of people involved in delivering the service.
Necessary calculations for the following were designed:
- Consumption and nominal carbon footprint of AIsaac and the SOC workforce
- Actual Carbon footprint of AIsaac and the SOC workforce
- Carbon credits required to achieve Carbon Neutrality based on actuals
- Identification of global/recognized reliable Carbon Credits provider for buying Carbon Credits form to achieve Carbon Neutrality
2. Data Gathering
Data pertaining to AIsaac and the SOC workforce was gathered by using the following approach:
a. AIsaac perspective:
- An inventory of virtual machines by global regions (India, Europe & North America), storage and usage levels was created.
- For each Virtual Machine, type, quantity, total vCPUs used, total RAM used, and average utilization factor were determined. The average utilization factor was derived from the Azure Usage Report.
- Inventory of storage (data at rest) was determined by using the Azure Usage Report on Data Stored.
- Inventory of flow of data (data in motion) was determined by using the Azure Usage Report-Bandwidth-Data In and Data-Out.
b. People perspective:
The number of people working by country on the AIsaac platform and in SOC Operations was tabulated.
3. Assessment & Calculations
This phase consisted of doing the various calculations needed for computing the actual CO2 emissions and costs.
a. Consumption Calculations
Carbon Footprint calculation for the technology part is directly dependent on the power consumption.
In the case of Compute (Virtual Machines), power consumption is measured in kWh based on VM type, CPU Clock Speed, and RAM. To calculate the kWh the “Microsoft sustainability reference” was used. This documents the kWh consumption for each type of VM hosted in Azure. The actual consumption was calculated based on the resource utilization in the AIsaac environment.
In the case of data (both at rest and in motion), the kWh depends on the data storage utilization per virtual machine or the actual data transfer rate per virtual machine.
- For data at rest, we aggregated the data usage in TB per instance.
- For data in motion, we aggregated the GB per instance.
In the case of people, the carbon consumption depends on the number of people involved in different countries and was computed accordingly.
b. CO2 Footprint Nominal Calculation
The power consumption in kWh calculated for virtual machines, storage in TB computed for data at rest, data transfer rate in GB for data in motion, and people count by country were then converted into equivalent metric tons of CO2 greenhouse gas emissions (tCO2) to get the nominal values for the CO2 footprint of the MDR service. The following references were used for this purpose:
– To convert kWh to tCO2, we used carbon intensity rates based on “Atos emission profile export”.
– To convert TB of data at rest into kWh “Storage MS Cloud Carbon Study” was used which was then converted into tCO2.
– To convert GB of data in motion into kWh “GB Lean ICT Report Shift Project 2019” was used which was then converted into tCO2.
– To convert headcount into tCO2 “Atos NetZERO Team reference Documentation” was used.
The calculations were then tabulated as presented in the following tables:
c. CO2 Credits Needed
Based on the nominal Carbon footprint calculation for the AIsaac platform, the actuals were derived based on platform lifecycle and the cloud computing environment used. The actual CO2 footprint turned out to be lower due to the impact of GREEN Energy used in Azure cloud computing.
4. Purchase of Carbon Credits
There was a residual incompressible part of the carbon footprint related to the overall MDR service as presented in the tables above. Hence, a need for purchasing Carbon Credits arose to neutralize this residual Carbon footprint. Carbon credits were then purchased on a yearly basis to achieve the Carbon Neutral status of the Cloud MDR Service.
Achieving Carbon Neutral status for the Cloud MDR Service was not a standalone project but rather a journey. Multiple strategies and continual improvements programs are in place to maintain the Carbon Neutral status of the MDR service. We will endeavor to do the following:
• Continue to participate in the NetZero engagement and activities at the organization level
• Continue to host the AIsaac platform in a cloud environment using green power
• Continue to optimize the technology to achieve/handle more with lesser technology resources (compute & storage)
• Continue to focus on automation to reduce the people/employees required to deliver the MDR service.
About the authors
Group Manager – AIsaac Architecture
Amirthamurugaraj Sargurunathan – Group Manager – AIsaac Architecture
Amirthamurugaraj Sargurunathan is a seasoned IT professional with over 25 years of diverse experience in roles spanning IT management, software development, product management, and information security. Currently, he leads the AIsaac Architecture Team within the Global CTO Team, where he focuses on architecting complex enterprise IT environments for MDR service delivery using AIsaac. With a strong background in software development and product evangelism, he has been a key contributor to Paladion/Atos/Eviden for over 15 years, handling various responsibilities including software development, product design, and global implementation for AIsaac-based MDR services. He holds a degree in Engineering with a specialization in Electronics & Communication.
Global Head – AIsaac Product Engineering
Ravi Raman – Global Head – AIsaac Product Engineering
Ravi Raman has over 30 years of industry experience, including many years in product engineering and management. During this span, he has worked with leading organizations in India and U.S. and managed a large number of enterprise class products. He has developed products in a variety of domains, including security, banking, financial services, insurance, health care and e-commerce. It has been a disparate mix of technologies – .NET, Java, Open Source and BigData. Ravi Raman has also played a critical role on the engineering the Alsaac platform since the beginning of its development.