AI Projects For EHS: Precursors To Success

AI Projects For EHS: Precursors To Success

There is no question that AI has hit the mainstream within the corporate world, largely thanks to recent step-changes in its sophistication, and the release of generative AI products, such as Open AI’s ChatGPT. According to a 2022 study by the UK government, 68% of large organizations use AI in some capacity, with that percentage forecast to increase in the coming years.

Although EHS is generally not pioneering AI in organizations – that’s usually left to finance, customer service or marketing functions – there is growing demand for its use. Through buyer and vendor conversations, as well as our global survey data, Verdantix sees consistent evidence of EHS functions’ desire for intelligent systems that can simultaneously deliver efficiency and performance benefits. Consider our 2023 survey of 301 EHS decision-makers, which found that in aggregate, 42% of firms wish to increase their usage of, or start to pilot, AI for EHS in 2024.

With the combination of strong market demand for intelligent EHS solutions, a vendor landscape with an abundance of embedded EHS AI use cases, and 77% of EHS budgets set to rise in 2024, the stage looks set for a significant increase in AI adoption. However, before diving into the arena headfirst, buyers should be mindful that successful AI projects, particularly novel ones, can require substantial groundwork.

Verdantix engaged in conversations with AI-focused EHS software vendors, and customers who have deployed custom EHS AI projects, to understand crucial considerations for the delivery of accurate outputs. What is the main takeaway? In one word – data. Specifically, firms should focus on aligning their organization-wide data architecture. This entails developing a standardized global labelling taxonomy, establishing consistent rules on how data and metrics are defined and calculated, and setting minimum data integrity standards. At present, most large organizations have disparate systems for EHS across commercial point solutions, Excel and paper-based processes. This generates a multitude of issues, such as siloed data sets, non-digital or incorrectly formatted data, inconsistent data collection methodologies, and below-par data quality. While resolving these issues may not rise to the top of the priority list in 2023, if left unaddressed, poor data quality can become a limiting factor for EHS performance.  

Many software vendors consider EHS data health-checks to be a positive side-effect of AI projects, forcing firms to critically assess the quality of their EHS data. Through the discovery phase of AI projects, firms can work collaboratively with software vendors and service providers to develop a data improvement plan. Additionally, project sponsors should engage their internal IT function to establish and enforce EHS data governance programmes – or even use AI to improve data fidelity by identifying anomalous, duplicated, missing or erroneous data.

To learn more about the precursors to EHS AI project success, as well as the abundance of EHS AI use cases on the market, read Verdantix Strategic Focus: AI And The Revolution Of EHS Compliance.