UT PhD candidate Thomas Leigh Hackett has devised a way to improve the air velocity measurements of a so-called flow sensor by as much as 511%. This could lead to better sensors for horticulture and better airflow in data centers.
Unlike traditional flow sensors, which are designed to measure just one parameter, Hackett's method captures a range of data using a single MEMS (microelectromechanical systems) chip. The sensor can help optimize conditions in vertical farms and manage airflow in data centers, where visibility into the entire environment is critical.
"With just one flow sensor, we can now capture much more information," says Thomas. "That means we can reduce the amount of expensive hardware and invest more in advanced analytics. Our sensor isn't perfect in the conventional sense—it's more sensitive to those external factors. Instead of just capturing flow data, it's also picking up temperature, humidity, and other environmental cues. We see this as an opportunity, not a hindrance. It allows us to capture more information using one sensor."
All on one chip
In addition to reinventing sensing techniques, Thomas is also trying to improve the way flow sensors are made. Most flow sensors are manufactured using bulk micromachining — an extensive and chemically demanding process. With a CMOS-based design, however, both the sensor and the data processing are made on the same silicon chip.
"Now the brain power is on the sensor itself," Thomas says. Traditional sensors require two separate chips: one to capture the signal and one to interpret it. The two are connected by tiny wires, adding steps and costs to the process. But with CMOS, the sensor and the computer chip are on the same piece of silicon. So the sensor (on the top layers of the chip) captures the data and the computer (on the bottom layers) interprets it — all on a single chip. "This also halves the manufacturing cost because we don't need the MEMS foundry: it's all done in the same facility," Thomas says.
Smart sensors for smart horticulture
The potential is huge. In greenhouses and vertical farms, for example, airflow is essential to plant health. It reduces mold, disease and other risks caused by stagnant, moist air. "Our sensors can detect air movement as low as 0 to 1 meter per second. By placing these sensors around plants, farmers can track airflow with astonishing accuracy." They can also track temperature and humidity to create optimal growing conditions that promote higher crop yields.
Data center cooling
Data centers, with long rows of servers running 24/7, can also benefit from these sensors. To prevent overheating, data centers use either air cooling or liquid cooling, with air cooling being the most common and less expensive option. Air cooling relies on fans blowing cool air over the servers. But in these large spaces, airflow isn't always smooth. Some areas don't get enough air movement, creating what are known as dead zones. Stagnant air in these areas can cause heat to build up. This can slow down servers, damage hardware, and in extreme cases, even start a fire.
"We're working on having our sensors measure airflow in real-time. By placing these sensors throughout a data center, you can detect dead zones so you can adjust fans or cooling setups to move air more evenly," Thomas says.
IEEE sensors conference
Thomas recently presented his research at the IEEE Conference in Japan. 'I heard a top speaker say that Twente is a world leader in flow sensors and MEMS technology. It was surreal to see our work cited.'
Looking back on his research, he adds: "It is exciting to do a PhD at a university that is world-leading in this field." He hopes to continue his research in the areas of CMOS and MEMS integration, scale production, and other potential applications for this type of flow sensor in various industries.
Source: University of Twente