Stihl: automating tedious compliance tasks
You might know Stihl from their chainsaws. To meet new regulations, they needed to create thousands of repairability labels for their products. The old way of doing things was a manual process, involving translations and Photoshop, that took a solid 15-20 minutes for every single label, which quickly adds up.
Before AI, this repetitive task added up to around 80 kWh of laptop use each year, which translates to 18 kg of CO₂.
After AI, things look very different. An AI pipeline now creates the labels in a matter of seconds. All their experts have to do is a final quality check, based on their extensive experience. This new, automated process uses just 2.4 kWh per year.
The 97% energy saving is impressive, but the real win is for their people. The compliance team is no longer stuck doing tedious, repetitive work. Instead, they can focus on what they're actually good at: high-value tasks like regulatory interpretation and quality control.
Logistics: optimising machine use
At a major international logistics hub, the cargo scanners were scheduled manually. This was a classic case of inefficiency: on some days, the expensive machines would sit idle for hours, while on others, they were completely overused and caused bottlenecks.
Each scanner used to run for about 12 hours a day, whether it was needed or not. That adds up to a consumption of 3,000 kWh per year, for each machine. After AI, a predictive scheduling model makes sure the machines only run when there's actual work to be done.
The impact is straightforward and immediately noticeable: the daily runtime dropped by four hours per scanner. That's a saving of 1,000 kWh and 230 kg of CO₂ per scanner, every single year.
The BOB project: freeing recruiters from paperwork
Recruitment should be about one thing: people. But in reality, a huge amount of time is spent on administrative work, like reformatting CVs for client proposals. On average, this tedious process took several hours for every single proposal.
All that manual work added up to around 360 kWh of laptop use annually. To fix this, we built BOB, an internal AI tool for De Cronos Groep. It now handles the structuring and formatting in just a few minutes, bringing the total energy use for the same work down to just 5.5 kWh per year: a decrease of 98%!
But the real impact of BOB goes far beyond energy savings. It addresses two core principles of good governance and social responsibility:
- The tool is designed purely for formatting and structuring. It does not and cannot make content changes to a CV. This guarantees that the information remains trustworthy and accurate throughout the process.
- By standardising how candidates are presented and scored based on purely factual criteria, BOB helps to remove unconscious bias from the initial selection process. A crucial step towards fairer recruitment, supporting diversity and inclusion goals.
Finally, the change fundamentally improves the work itself. Recruiters are no longer wrestling with documents, so they’re free to spend their time doing what actually matters: interviewing and connecting with candidates.
Beyond the kilowatts: the human and governance side
As these examples show, a smart AI implementation does more than just cut down on energy use. The benefits of AI in ESG touch on the social and governance aspects as well.
When you automate monotonous and repetitive work, you're not just saving time but also improving people's jobs. It frees up your employees to focus on the creative, strategic, and human-centric tasks they were actually hired for. The result is better job satisfaction and a more engaged workforce.
Governance is another key aspect, since trust in AI is built on one thing: responsibility. That’s why we design systems with transparency and auditability from the very beginning. For us, this means always keeping a human in the loop to make the final decision, using highly specific data to reduce bias, and making sure everything complies with regulations like the EU AI Act.
Where to start with AI in your organisation
Our approach is not to sell AI for the sake of it. In fact, many business processes can be improved without it. We always start with a simple question: where does it hurt?
By identifying a specific and wasteful process in your organisation, we can measure its current footprint and determine if AI is the right tool for the job. If it is, we can help you pilot a solution responsibly, report the results transparently, and scale what works.
