How AI Is Transforming Analytical Science Workflows
Discover how scientists are using AI to automate analytical workflows and transform laboratories.
Complete the form below to unlock access to ALL audio articles.
Artificial intelligence (AI) has rapidly shifted from an emerging trend to an indispensable force in analytical science. By automating repetitive tasks, enhancing data accuracy and unlocking predictive insights, AI is already reshaping how laboratories operate – and this is only the beginning.
From streamlining data analysis to optimizing instrument management and experimental design, AI is becoming deeply embedded in analytical workflows. To explore its growing impact, Technology Networks asked experts at the 2025 American Society for Mass Spectrometry (ASMS) conference a single question: How can AI support analytical science workflows, both today and in the years ahead?
Automating data processing to save time
“AI has become mainstream,” said Jim Gearing, associate vice president of marketing in Agilent's Gas Phase Separations Division. “About two years ago, with the introduction of ChatGPT, everybody started to learn what AI was about.”
Agilent’s acquisition of Virtual Control marked a key step in bringing AI and machine learning (ML) into its portfolio. Its first application targeted phthalate analysis using gas chromatograph-mass spectrometry.
“What it [the analysis] does is use AI and ML to do peak integration,” explained Gearing. “An AI/ML model can be trained with hundreds or thousands – maybe tens of thousands – of analyses of everything a laboratory might see: ‘this is good,’ ‘this is bad’ and everything in between.”
Once trained, these models can automatically process future samples, providing peak integration and identification in real time. “It significantly shortens the user’s time for analysis,” Gearing said. “AI’s biggest play is increasing throughput and shortening time to answer, so people in the lab can spend less time in front of a computer and more time on valuable tasks that matter to their enterprise.”
Monitoring instrument health and predicting maintenance
AI’s value also lies in improving instrument performance management.
“AI and ML have become more and more important,” said Dr. Thomas Moehring, senior director of OMICS applications and managing director at Thermo Fisher Scientific. “We already apply it to monitor instruments and predict when they need maintenance, or how performance is tracking over time.”
This predictive capability reduces downtime and helps labs run more efficiently. But, like Gearing, Moehring also highlights AI’s role in managing data volume. “With platforms like the Orbitrap Astral, the amount of data generated is immense,” he explained. “How do we ensure all this data is processed and analyzed in a meaningful way?”
The answer lies in AI-driven data workflows. “We’ve implemented ML algorithms like Ferris CHYMERYS into our Proteome Discoverer platform,” Moehring added. “It’s about making better use of the data we generate – and this will continue to be central to our future development.”
Bridging data acquisition and analysis
AI is also poised to unify data acquisition and interpretation into a seamless feedback loop.
“MS [mass spectrometry] data is highly informative, but there are also a lot of artifacts you have to exclude,” said Dr. Jose Castro-Perez, vice president of product management at SCIEX. “Having an AI system that can help make those decisions on the fly, while communicating with the instrument – saying, ‘here’s an experiment you need to do next’ – would be transformative.”
Such integration could allow AI to dynamically adjust experiments as data is collected, reducing wasted runs and improving outcomes. “It’s about interpreting the data and connecting acquisition with processing,” Castro-Perez said. “That leads to faster throughput, speed to data and better decision-making in real time.”
Castro-Perez also points to AI’s potential in building predictive biological models. “Think about models for different organisms or cell types,” he said. “You could deeply characterize phenotypes and then use those to build even more predictive models that lead to better therapies in the future.”
A future of seamless, intelligent workflows
Ultimately, AI’s role in analytical science spans every part of the workflow: guiding experiments, improving data quality, accelerating analysis and even connecting instrument maintenance to lab scheduling.
“As I see it,” Castro-Perez summarized, “it’s about bringing instrument acquisition and setup together with data processing and then networking that with other data inputs. When all those pieces are connected, AI has the power to transform how we do science.”
With AI advancing rapidly and instrument vendors already embedding ML capabilities into their platforms, the analytical lab of the future may be one where scientists spend less time on repetitive tasks – and more time focusing on discovery and innovation.
