1660601175656 Ct20weblogo550

Water analyzer relies on robotics and the cloud

April 9, 2021

Ketos Inc. has developed a water quality sampling and testing system, Ketos Shield, which employs multi-axial robotics and embedded processors to measure 26 parameters, provide real-time insights, and use Internet links to relay results to a cloud-based analytics service—so users don't have to visit their hardware and facilities so often. Its robotics deliver samples to stacked, vertically integrated sensing and processing components inside its enclosure. The company adds that modular, bench-top, microwave oven-sized and autonomous Ketos Shield also eliminates the need for static water testing, constant cleaning or manual calibration.

"Handheld probes and analyzers measuring pH, conductivity, chlorine and other parameters have gained online capabilities in recent years, but they still use up consumable resources and require users to visit their physical equipment and operations every few days and deal with data proliferating from multiple sources," says Meena Sankaran, founder and CEO at Ketos. "We've come up with a way to shift the water quality operations mindset to a way to collect and analyze more data. We asked users to list all the parameters they wanted, including heavy metals, took their EPA target limits, and developed a system to measure all of them with 95% confidence."

Ketos Shield is reportedly ideal for high-frequency water sampling applications, such as ground and surface source-water monitoring, industrial process water, reservoirs, processed fracking water and drinking-water quality. It's available on a subscription basis; sends raw data from its applications via Ethernet and wireless networking protocols to its cloud-based analytics platform; and can store and integrate year of static data to provide a unified view of operations.

Because it was designed and built in-house, Ketos Shield's microcontroller employs its own application program interface (API) to interact with all types of external data processing systems and services. It can give users real-time alerts, EPA or customized threshold-based diagnostics, custom reports, historical trends and actionable insights by continuously analyzing water data with machine learning (ML) principles. This not only provides early warnings about water quality, but also lets predictive modeling optimize process controls of chemical feeds, groundwater contamination and water reuse.

"If a user is running numerous pumps and regulators on Siemens Mindsphere or another platform, we can integrate with all of them," explains Sankaran. "Ketos API can pass data to whoever needs it. We also partner with OSIsoft, so PI System users can access Ketos Shield through the API, or send data points to it to help develop predictive models."