Become an Agile Environmental Lab: Part III, Artificial Intelligence

In laboratories around the world, artificial intelligence (AI) is being paired with robotic and automated technologies to expand the scope of science by harnessing machine learning (ML) and hypothesis generation to improve experimental design and data analysis.

The environmental testing industry has suffered recently a period of consolidation, intense price competition, and declining profitability-and labs should adopt AI technologies that can help alleviate these pain points.

Between 2016 and 2018 in the U.S., there was a significant decrease in laboratory technicians – a statistic that is reflected in the environmental testing industry. However, automation, AI, and ML can overcome this by taking over parts of the workflow.

AI in environmental testing laboratories may introduce some challenges in the workplace. Similar to automation and the cloud, there may be substantial upfront costs and training may be required for staff. Additionally, data silos can limit the abilities of AI system, but data silos can be removed if a centralized data storage system is created and combined with a management system.

Another limitation is the ethical considerations; the UK’s science academy raised the concern that ML will only benefit a select few, and that to make all advantages inclusive, ML required supervision from scientists. But when programmed correctly, AI produces fewer errors and biases than humans. This has resulted in successful implementation into environmental testing protocols; AI has been used in mineralogical composition prediction, which, combined with simple infrared spectrophotometry testing for chemical and physical properties, can reduce the testing to a fraction of the cost

To continue learning about how you can futureproof your lab, download the complete white paper now.