As a company dedicated to helping industrial manufacturers become more sustainable through process optimization, we also track and seek to reduce our own impact on the environment.
This Earth Day, we teamed up with French startup Carbometrix to measure our own carbon footprint and learn what we could do to reduce it.
As many as 99% of companies have no idea what their carbon footprint is, according to Carbometrix, and 80% of the carbon footprint of a company like ours is often indirect. That makes measuring it — and consequently taking action to reduce it — challenging.
We provided Carbometrix information about the IT infrastructure we use, our office spaces, and even our travel behavior. After about a week, they got back to us with a report summarizing our total emissions for 2021.
Last year, Fero Labs emitted 76 tons of CO2, according to the report — equivalent to the average yearly footprint of six Americans. Most of our emissions came from our office, with electricity consumption accounting for 42% of the total.
Compared to our peers in the AI and ML space, we outperformed on IT because our white-box machine learning technology is also energy-efficient.
On a call, their team explained that we use very little energy compared to other ML companies, because we’re using much more efficient algorithms compared to the typical application of black-box machine learning methods, like deep learning. This reduces our energy consumption significantly.
The report also revealed how our results stacked up to the GHG Protocol Corporate Accounting and Reporting Standard. This standard breaks down emissions into three categories: Scope 1, direct emissions; Scope 2, indirect emissions from power generation; and Scope 3, all other indirect emissions.
Looking at it this way, Scope 3 accounted for most of Fero’s emissions, which is typical for many companies in our sector, according to Carbometrix. This is where we are focused on reductions.
What we learned
Overall, the experience of working with Carbometrix was very smooth. Carbon accounting is really difficult, and Carbometrix made it easy for us to get an estimate of our emissions.
When it comes to concretely reducing emissions, the report also provided recommendations, with a focus on Scope 3 reductions. While we don’t own the buildings where we rent office space, we can communicate our desire to control our heating more precisely; similarly, air conditioning is in our control, and we will track and reduce consumption in the warmer months. In terms of electricity, we plan to communicate our desire to our landlords to switch to renewable energy providers, for which there are many options in New York City.
This process highlights the crucial role of data and analytics in achieving sustainability goals. Even though our emissions were reasonably small in scope, relative to the factories that use our software, it was helpful to track and measure them. In a factory where thousands of data points are generated every second, it’s even more vital to implement software that can take them all into account, helping manufacturers set their priorities and achieve their corporate sustainability goals.