Modern lab monitoring systems have a significant impact on how labs operate by providing real-time data that can improve safety, compliance, and efficiency. Combining a wide variety of sensor data with powerful AI/ML-driven software applications and dashboards provides lab managers with incredible insight into the operation and utilization of equipment and instruments. 

For a broader perspective on the forefront of lab monitoring, we spoke with Sridhar Iyengar (SI), founder and chief strategy & technology officer of Elemental Machines.

Get training in Asset Management and earn CEUs.

One of over 25 IACET-accredited courses in the Academy.

Asset Management course

What are some challenges that labs face when using traditional equipment monitoring approaches?

SI: One of the key challenges is regulatory inertia, which often necessitates revalidation, rewriting of standard operating procedures (SOPs), and re-training of staff. These changes cost labs both time and money and ultimately lost productivity. Another challenge is how often equipment gets moved and relocated within labs. Having a fully wireless monitoring system that can easily move with the equipment is a tremendous benefit. In the end, labs need the data served to them where and when they need it. Given the longevity of scientific instrumentation, labs are making purchasing decisions about the right platform, not just for today, but for the next 10+ years.

Are there ways to aggregate monitoring data into convenient applications or tools to improve the efficiency of making decisions in the moment?

SI: Lab monitoring data is more valuable when it is combined with operational and financial data, since ultimately the goals of any organization are overall efficiency and productivity. There is great value in aggregating monitoring data with the company’s other data sources using data lake platforms and business intelligence (BI) tools (e.g., Snowflake, Tableau, PowerBI, etc.) to gain insights for purchasing and maintenance decisions. Examining usage and utilization data can reveal discrepancies between scheduled and actual equipment use to improve resource allocation and space utilization, as well as preventative maintenance planning.

It is important that the monitoring data is generated in a format containing the appropriate metadata that enables it to be easily aggregated in larger BI systems. Some best practices include consistent naming conventions, enforcing categorization, involving developers in designing data flows, and establishing SOPs for data handling prior to aggregating lab monitoring data with a larger data lake.

How would most labs use the ability to configure monitoring software? Why would this be valuable?

SI: The key to configuration is to enable the lab to customize alert thresholds, recipients, how alerts are delivered, and actions for different situations or sites. The ways that alerts are configured might differ greatly between labs associated with manufacturing and quality control versus labs involved in research and development (R&D). Manufacturing might configure notifications based on different shift times, while R&D might focus on equipment moves and locations. Another alert configuration we’ve seen is for alerts to be sent to another software system via APIs that then trigger additional actions in those other systems.

How can AI tools help labs improve how they monitor equipment?

SI: Traditionally, lab monitoring data was used retrospectively to understand that some alert threshold was crossed. With more powerful AI software tools, historical data can be used to predict future equipment behavior and identify relationships not immediately apparent. This can be very effective in determining root causes. To be predictive, the AI models require large amounts of high-quality data and allow for proactive instead of reactive maintenance. Aggregated customer data is used to train some AI models. In addition to the rise of large language models (LLMs), we are seeing more AI agents being developed to analyze data for specific applications, such as predicting equipment failures, recommending optimal operating settings, etc.

What additional data is available to help labs improve instrument use, optimize efficiency, and reduce costs?

SI: The biggest costs in most labs are personnel, capital expenditure, and maintenance costs. Lab monitoring data can drive improvements in all of these key areas. The data can teach labs how people interact with equipment, and anonymized personnel movement data can be used to optimize lab layouts and workflow. Combining operational and financial data creates opportunities for comprehensive cost optimization, including aiding in decisions of when to retire equipment or whether to renew a maintenance contract for equipment that is rarely used. We've seen examples of companies recovering enough space to eliminate planned expansions, which translated to substantial rent savings.

With labs increasingly relying on cloud-based data platforms, how can they ensure simple yet secure access to critical data?

SI: Labs need to secure cloud data access through staff training, network security measures (firewalls and whitelisting of their sensor devices), and by choosing reputable providers with robust security practices and disaster recovery plans. Labs can work with their providers to reduce the number of points of vulnerability to attack, and to vet the providers’ security activities. 

How important is customization of monitoring applications? How would a lab use customization to improve its operations?

SI: While customization of a basic monitoring platform is very useful (i.e. in terms of customizing alerting rules and recipients), where we have seen the most impact is when a platform offers customization of data analytics and visualization. With the proper data and visualization platform, labs can customize how they interact with the data by using flexible querying of their data to find hidden relationships that can reveal inefficiencies in their operations. This empowers lab managers to perform their own in-depth interrogation and analysis of the monitoring data and see their data the way they want to see it (and virtually every lab will want to see their data in different ways). We have seen lab managers compare utilization data (how much an instrument gets used) to maintenance records to better plan their preventive maintenance cycles to minimize unplanned downtime.

How can labs improve their approach to real-time data monitoring to extract more value?

SI: The key is to focus on actionable insights from real-time data that can show financial ROI. Prioritize real-time monitoring for events requiring immediate action rather than retroactive monitoring for less time-sensitive data analysis. The highest value comes from understanding the actions to be taken in relation to the collected data before committing to real-time monitoring costs. Especially in this economy, it’s important to ensure that any investment in a monitoring platform can be justified by having multiple stakeholders receive value from the data. 

For example, lab managers obviously benefit from real-time alerting, but with the proper analytics and visualization capabilities, the procurement team can make better purchasing decisions based on usage data, the facilities managers can see if HVAC fluctuations are causing more calibration issues for certain equipment, and scientists can review equipment data to ensure that set points on equipment were indeed stable. 

What does the future hold for lab monitoring? What is the next exciting development?

SI: Future advancements will center around advanced data visualization, like 3D immersive visualizations for remote access, and agent-based AI tools, enabling more intuitive querying and analysis of large datasets, like tools like Google’s NotebookLM. It is exciting to think about how these AI tools can accelerate more powerful, natural inquiry of the data.

Lab monitoring data can be used in powerful ways to extract more learning from lab operations. This learning will occur across all the applications of lab monitoring, but may be most valuable when the data is aggregated with other operational and financial data. These improvements are being driven by powerful AI tools that make querying and visualizing the data more intuitive and faster.

A serial entrepreneur, Sridhar Iyengar has significantly impacted the connected medical devices and wearables industry. Prior to his role at Elemental Machines, he co-founded Misfit, which was acquired by Fossil in 2015, and AgaMatrix, a pioneer in developing medical devices integrating with smartphones. Sridhar’s strategic focus is underscored by his rich patent portfolio and his academic pedigree from Cambridge University, where he was a Marshall Scholar. His vision for Elemental Machines reflects his passion for innovative, data-driven solutions.