Blog

Implementing AI for sustainability

Jonna Saarinen Senior Service Designer, Solita

Published 30 Jan 2025

Reading time 5 min

In my previous blog post, I explored how to establish a solid foundation before implementing AI solutions. Following design thinking principles – which emphasise user needs, exploration, and iteration. I covered these three key phases:

  • Defining goals and challenges
  • Identifying key opportunities
  • Assessing data quality and availability

Read the previous blog post: Getting started with AI for sustainability

Now, let’s explore how to move from planning to action. I’ll focus on three phases that will help you integrate AI into your sustainability efforts:

  • Exploring available AI solutions and finding the right fit
  • Evaluating and prioritising solutions based on value and feasibility
  • Developing and scaling solutions

1. Explore solution options

AI offers a range of solutions to various sustainability needs:

Predictive analytics tools: Major cloud platforms like Microsoft Azure, AWS, and Google Cloud offer both general analytics and specialised sustainability solutions. For example, Microsoft Sustainability Manager helps track environmental impact, while Google Cloud’s Carbon Footprint focuses on emissions monitoring.

Generative AI platforms: Tools like Claude or ChatGPT can analyse vast amounts of data to generate valuable insights and recommendations. For example, they can help analyse regulations to suggest tailored sustainability initiatives, or assess supply chain data to identify sustainability risks and opportunities.

Specialised sustainability software combine AI with domain expertise:

  • ESG management and reporting: Tools like Workiva simplify ESG reporting through integrated AI capabilities. By automating analysis and documentation tasks, teams can quickly create and edit content and generate insights. For example, the platform helps teams create an ESG double materiality assessment or draft a climate impact analysis.
  • Carbon accounting: Solutions like Persefoni help companies track and reduce their carbon emissions using AI. Persefoni converts your business data – like electricity usage or travel – into carbon footprint numbers. It can also spot unusual patterns and provide technical support when needed. This makes carbon accounting faster and more accurate, letting teams focus on actually reducing emissions rather than just measuring them.
  • Life cycle assessment (LCA): Tools like Makersite apply AI to analyse product environmental impacts throughout their lifecycle. This helps companies understand and optimise their products’ environmental footprint from raw materials to end-of-life. For example, Microsoft used AI-powered LCA to spot improvement areas in their supply chain, cutting the Surface Pro 10’s carbon footprint by 28% in two years.
  • Supply chain risk management: Tools like Prewave use AI to help companies understand their entire supply chain, from direct suppliers to their partners’ partners. The system combines public information like news with private supply chain records and shared customer insights. This helps companies spot potential issues early, ensure compliance, and keep operations running smoothly even if unexpected challenges arise. For example, Kärcher, a cleaning equipment manufacturer, used Prewave’s AI to reduce their compliance monitoring time by 40x.

For organisations with unique requirements, custom solutions through enterprise platforms offer maximum flexibility. This means building tailored applications using platforms like Microsoft Azure, AWS, or Google Cloud that can combine multiple AI capabilities.

Often, a hybrid approach—combining ready-made tools with custom solutions—can deliver the most comprehensive results. For example, you might use specialised carbon accounting software for emissions tracking while developing custom AI models for your unique supply chain optimisation needs.

Example from Persefoni: Converting business data into carbon footprint numbers. Here, AI automatically matches different types of business activities with their appropriate emission factors to calculate the carbon impact.

AI example from Persefoni

2. Evaluate and prioritise solutions

After exploring different options, carefully evaluate each solution against these key factors:

Strategic alignment: Does the solution support your key sustainability goals? For example, if carbon reduction is your focus, look for tools with robust carbon tracking and reduction planning capabilities.

Stakeholder value: Consider how AI solutions can benefit your key stakeholders. Leadership gains strategic insights and sustainability experts get enhanced analytical capabilities. Suppliers benefit from improved collaboration, investors receive better ESG reporting, and customers gain clearer sustainability information.

Implementation feasibility:

  • Data quality and availability: Is your data accurate and complete? Can you access it easily? Your AI tools will need quality data to deliver reliable results.
  • System integration: How well will the solution fit with your current IT infrastructure?
  • Resource needs: What technology, personnel, and budget will you need?
  • Capabilities: What skills does your team have, and what training might they need?
  • Timeline planning: What’s a realistic schedule for setup, training, and deployment?

Cost-benefit analysis: Look at the complete picture when evaluating solutions. Weigh implementation costs against both business and sustainability benefits. Look for solutions that deliver both quick improvements and lasting value.

Regulatory compliance: Choose solutions that meet current regulations and can adapt to new requirements as they emerge.

Vendor capabilities: Choose providers with proven technology, innovation capacity, and industry expertise. Verify their financial stability and development roadmap align with your needs.

3. Develop and scale solutions

Starting your AI journey doesn’t have to be overwhelming. Begin with manageable steps that align with your capabilities and gradually expand as you gain experience.

Start with focused pilots. You can pick a specific challenge—like optimising your supply chain—to test out the AI solution. Pilots help you gather insights, validate approaches, and learn valuable lessons before broader implementation.

Measure and iterate. Track your success with clear KPIs, whether it’s cutting emissions, saving costs, or improving resource efficiency. Use these metrics to demonstrate value and guide development. Keep adjusting based on feedback and early results.

Scale solutions. After successful pilots, expand proven solutions across your organisation. You can start with similar operations – for instance, if AI helped one business unit automate its emissions reporting, expand the solution to units with similar reporting needs. Then gradually expand to other areas, monitoring performance against your sustainability goals.

Ensure governance. Establish strong oversight for AI deployment. Make sure your data is accurate, secure, and well-managed. Having clear oversight ensures your AI outputs are reliable.

Develop competences. Equip your team with the skills needed to effectively use AI tools. Ongoing training fosters innovation and makes it easier for everyone to adopt AI-driven practices.

Implementing AI for sustainability takes careful planning, but the rewards are worth the effort. Companies that begin this journey today will be better positioned to meet growing stakeholder expectations and environmental challenges tomorrow.

  1. Business
  2. Tech