One-Third of Sustainability AI Implementations Are Driven by Just 10% of Companies
AI ROI comes in various forms, and sustainability is one of them. Our research found that 10% of the companies in our data set are responsible for almost a third of AI implementations relating to sustainability in supply chain. We delve into what PepsiCo, part of that 10%, is doing.
The Data
-
1
10% of companies in our data set of 255 are responsible for 31% of sustainability use cases in our AI Hub.
-
2
This 10% of companies (who we refer to as AI leaders in this Data Insight) are talking about, hiring for, and implementing AI the most.
-
3
The most common category of AI use in sustainability is waste reduction (24% of use cases), followed by energy management and product development (both 14%).
AI Leaders Charge Ahead on Sustainability
According to PwC UK, the use of AI for environmental applications could mean significant productivity improvements between now and 2030. It could contribute up to $5.2 trillion USD to the global economy (a 4.4% increase), reduce worldwide GHG emissions by 4%, and create 38.2 million net new jobs for skilled workers across the global economy. Getting to this point, however, involves reducing the negative environmental impact inherent in all computing tools – namely, the amount of energy and water consumption required to run them – and making use of AI for sustainability.
Just 5% of use cases in The AI Hub relate to sustainability, with other priorities like speed, efficiency, and cost edging out this increasingly urgent consideration. We zoomed in on the use cases that made up that 5%, and our analysis found that 31% of those use cases come from just 10% of the companies in our data set.
When we expanded to the top 20% of leaders, that 31% of use cases only rose by 7%. This indicates a wide gap between the top 10% and the top 20% of AI leaders, showing that AI leaders are investing significantly more in building out AI for sustainability.
PepsiCo Revolutionizes Recycling with AI
PepsiCo, part of this top 10%, has been using AI to reduce plastic waste since 2021. With the goal of optimizing recyclability through systemic changes, the beverage giant has implemented AI to more reliably identify and sort plastic waste, which includes improving sortation systems.
In 2022, PepsiCo, along with nine other brands, including Nestle, P&G, and Colgate, joined the “Perfect Sorting Consortium,” a project that involves an independent test and research center and universities, all with a goal of improving packaging waste sorting using AI. The two-year project is set to end this year – the final step being a successful test of the AI decision model in an industrial sorting plant, which can be used across Europe. Last year, the research center (NTCP) gave a progress update, revealing that the project had tested over 300 packaging materials, scouted multiple AI technologies, and that they were ready to enter a new phase: the realization of a demonstrator on AI-based packaging waste classification.
The Takeaway
Recognizing the power of AI to make a difference on your sustainability journey is crucial. Look at the sustainability initiatives at the top of your priority list, and consider how implementing AI, even on a small scale, could help expedite your progress. We suggest starting with the Zero100 AI Blueprint. The Blueprint is a self-assessment revealing key AI priority areas. Following this, explore use cases representing some of the tech solutions you may be able to implement (we recommend getting in touch with us to make use of our AI Hub).
To see a different data cut or to dig deeper into this topic, reach out to our Head of Research Analytics, Cody Stack, at Cody.Stack@zero100.com.
Methodology
Zero100’s proprietary data and analytics are a combined effort between our data scientists and research analysts. We provide data-first insights matched with our own research-backed points of view and bring this analysis to life via real-world case examples being led by supply chain practitioners today.
For this study, we analyzed ~7,500 data points across more than 30 different scoring components including earnings calls, LinkedIn job descriptions, and The AI Hub, disqualifying companies without publicly stated Scope 3 targets.