Nahrstedt and Team Create Sensing and Edge-Based Cyber-Infrastructure to Maintain Ultra-Clean Labs
6/3/2021 11:34:39 AM
When new technology and new products using sensitive materials are being designed and developed, it is essential that the environment is clean. One product that requires such an environment is semiconductor devices, which are used in a wide range of electronics, such as computer chips in laptops, personal computers and smartphones.
But think of how difficult it is to create a space where not even dust can get inside. It is with this aim that Klara Nahrstedt, the Grainger Distinguished Chair in Engineering Professor in Computer Science and Director of the Coordinated Science Laboratory, formed a team of Principal Investigators (PIs), Prof. John Dallesasse (HMNTL/ECE), Prof. Roy Campbell (CSL/CS), Dr. Kenton McHenry (NCSA) and Tracy Smith (Campus IT Infrastructure), to develop the project entitled ‘Sensory Network infrastructure for Scientific Lab Environments (SENSELET)’. The aim is to assist scientists in maintaining environmental control within cleanroom labs when fabricating new materials and semiconductor chips.
Nahrstedt answered some questions about the project, and how the team of PIs, students and cleanroom staff members are putting their computer science and cleanroom expertise to use to provide situational awareness of cleanrooms’ environmental status to lab users so that the materials and semiconductor device fabrication research can be conducted in ultra-clean labs.
Why are cleanrooms so important?
Cleanrooms, like those found in the Nick Holonyak Jr., Micro and Nanotechnology Laboratory (HMNTL) help prevent even a tiny amount of dust from interfering with experimental structures on a nanometric scale through constantly filtering the air throughout the room. The structures of these devices are often smaller than a single dust particle, so failing to filter all the dust out could lead to chip failures and significant delays in semiconductor research.
In addition to dust, we have to control other environmental parameters such as temperature and humidity. If these factors are not kept within certain parameters, they can result in the device having different dimensions from those required, causing the device to fail and, ultimately, the whole chip to fail.
How is SENSELET helping to ensure that cleanrooms are effective?
Steps can be taken to try and ensure that the cleanrooms operate properly. However, it is necessary to continuously monitor the room’s environment to discover whether the required conditions are being maintained. SENSELET is a cyber-infrastructure system that can collect and track data automatically, ensuring researchers can monitor and predict the environmental behavior around scientific equipment in any given cleanroom.
What are the main components of SENSELET?
SENSELET consists of three components. The first component is the environment monitoring module, which is composed of temperature, humidity, air-flow and water leakage sensors – and a small device we call SenseEdge to collect sensory data and push data to the remote private cloud, called SenseCloud. The second component is the data storage engine residing in the SenseCloud. This module runs a time series database which stores all the sensory data. The third component, residing in the SenseCloud, is the visualization and alert module, which provides visualization dashboards to lab administrators and sends out alerts when environmental data exceed a normal range.
What challenges did you face during the project?
The most difficult problem has been reliability, closely followed by maintenance and scale of diverse sensors. We need to build the SENSELET system as robust as possible to provide reliable cleanroom’s status updates 24/7. Hence, we designed a watchdog mechanism to manage the sensors.
Another issue we’ve faced involves the different types of sensors. We are working on software and hardware solutions to make sure SENSELET supports different sensor types and makes it simple to add or replace sensors. The final challenge is related to time-sensitive data analysis. We are working to shorten the time between the occurrence of an abnormal sensory event and the alert sent to a lab manager since an abnormal sensory event may indicate safety problems for users in cleanrooms. Our goal is to create a system that can predict future potential abnormal events based on the current situation.
What have been the key successes of the project?
We have used SENSELET here at Illinois and it has provided quantifiable data showing when the humidity and/or temperature controls are not within specifications. The system has also identified other problems in cleanrooms, for example, forgotten open fume hood after an experiment, warming up a cleanroom to an undesirable temperature. We created a dashboard to notify users of the current temperature and humidity deviations, measured inside the cleanroom, as well as historical data being tracked every minute. This data has proved useful in helping cleanroom lab managers and researchers adjust their management and scientific processes.
Besides PIs, the team working on this project includes Illinois CS graduate students Zhe Yang and Beitong Tian, Illinois ECE graduate students Patrick Su and Robert Kaufman, and HMNTL principal research engineer Mark McCollum.
This article was partially produced by Futurum, a magazine and online platform aimed at inspiring young people to follow a career in the sciences, research and technology. For more information, teaching resources, and course and career guides, see www.futurumcareers.com