Boavizta API : automated evaluation of environmental impacts of ICT services and equipments
Boavizta aims to mainstream the assessment and reduction of the environmental impacts of Information and Communication Technologies (ICT) in organizations by continuously improving its quality. For this objective, we created a working group, Boavizta, to deliver open-source tools, methods and open data. We are releasing an open-source API that allows the assessment of the environmental impacts of on-premise or cloud (AWS) equipment and infrastructure.
These impact assessments concern several criteria to date: Global Warming Potential (GWP), Primary Energy Consumption (PEC) and Abiotic Depletion Potential (ADP). We are currently considering two phases of the life cycle of these products and services: manufacturing and use.
For many reasons, assessing the environmental impacts of a digital service or system is a difficult task. Methodologies and data are heterogeneous, of varying quality and age, and cover diverse perimeters. They evolve rapidly at the pace of knowledge. Their integration and industrialization in tools and processes of organizations are therefore potentially costly without guaranteeing accuracy.
For these reasons, Boavizta has decided to propose an API (Rest) allowing to integrate and industrialize the process of measuring the environmental impacts of equipment, systems & digital services. We provide an interface to the data and methods allowing this evaluation, which the group will update and refine over time.
We centralize the results of our work on the state of the art of environmental impact methods and data. This API simplifies access to our work for direct evaluation of the impacts of an information system or service.
Examples of use cases that the API addresses:
I want to know the impact of my server based on its configuration and usage.
I want to know the impact of my cloud instance based on its type and its usage.
Examples of use cases we would like to address:
I want to know the impact of my information system from my inventory.
I want to know the impact of my service from its modeling.
I want to know the impact of my office equipment.
With this API, we want to create an ecosystem of services to facilitate the evaluation by integrating advanced methods and data collected and qualified by the working group.
This documentation explains how works and how to use the API.
Boavizta generously hosts a demonstration platform and an OpenAPI endpoint, for reasonnable use.
Users of the API are strongly encouraged to host their own version of the API, as we don’t guarantee the availability of the test platform, at all.
The code is available on Github.
Adapted to different levels of knowledge
In line with our commitment to mainstream the evaluation and the transition to action, our API adapts to the different levels of knowledge of users. It addresses the needs of experts as well as beginners, by allowing a more or less fine-tuned evaluation. The level of precision of the evaluation depends on the level of precision of the request and the data provided.
Transparency through Open Source and Open Data
We are convinced that transparency is one of the keys to a quality impact assessment, especially in a context where knowledge is evolving rapidly. Transparency requires, among other things, that the methods and data used are auditable. This is one of the reasons why we have chosen to offer an open-source code and open data allowing a critical review, external contributions and reuse by the greatest number. Moreover, the documentation of the methods used by the API and the traceability of its results are also levers of transparency and education to avoid the black box effect.
Multicriteria & Multi-steps
The impacts of digital technology on the environment are multiple and cannot be summed up simply as its contribution to climate change. These impacts occur at all stages of the equipment’s life cycle, from its manufacturing to its end of life, including of course its usage. At Boavizta, we are working towards a greater consideration of these elements, in a “Life Cycle Assessment philosophy”; this is why the API we are developing integrates the impacts of manufacturing and usage on three criteria: Global Warming Potential (GWP), Abiotic Resource Depletion (mineral & metal) (ADP) and Primary Energy Use (PE).
We have focused the MVP on the impacts of the manufacture and use of on-premise servers and cloud instances in line with the work of Boavizta and Benjamin Davy.
Today, we would like to move forward on several projects:
The addition of new components such as GPUs that would address the use case of mining, AI or high-performance computing
The addition of new equipment such as monitors, user terminals or the network
The addition of a system layer to measure the impact of multiple devices
The addition of a service layer to measure the impact of a digital service
The addition of end-of-life stages (waste, recycling, reuse) and transportation
The addition of new impact criteria (ionizing radiation, water impact, ecotoxicity, …)
If you want to help us, please contact Boavizta or participate directly on GitHub.
Cover image: “Mt Isa mines. One of the big holes adjacent to the city.” by denisbin is marked with CC BY-ND 2.0.