Artificial intelligence blog 2400x1350px

One step closer to Machine Learning in cleanroom technology

Category: Blog
Date: May 10, 2021

What is machine learning (ML) and more importantly, why would you need it? Machine learning is a subset of artificial intelligence (AI) that utilizes algorithms to sharpen the decisions businesses make. It can spot anomalies in how your equipment or assets are performing with advanced pattern recognition. Artificial intelligence can be seen as automating tasks that require human cognition, such as maintenance schedules for physical assets. Machine learning helps optimize the interference between humans and machines in an intelligent process that learns as it goes along. In other words, the predictive model continuously gets ‘tuned’. Many applications of machine learning are well-known: it is used on a daily basis for image and speech recognition, consumer marketing, medicinal diagnoses and so on. What it hasn’t been used for up until now, is for cleanroom technology.

The importance of HVAC in cleanrooms

Ever since the foundation of ABN Cleanroom Technology back in 1996, the company has unfolded itself as a design, construction, commissioning and validation company specialized in cleanroom activities.

Cleanrooms can be described as airtight rooms in which filter and air treatment technology keeps the concentration of contaminated particles in the air to a minimum. Often, a specific negative or positive pressure is created in these rooms in order to prevent outside air from entering the room, or inside air from escaping to the outside. In order to realise well-defined cleanroom conditions, all cleanrooms are equipped with several HVAC assets, such as chillers, fans, compressors, filters, humidifiers etc. For these HVAC assets, uptime is of crucial importance and failure of individual components or of the entire installation must be prevented.

Act when something treatens to happen

For every customer, a set of functional requirements is imposed within which a cleanroom must function 24/7 (= uptime). This set of requirements determines the range within which the parameters must remain to ensure optimum operation of the cleanroom (think of temperature, humidity, VOCs, particles…). As soon as one of these parameters exceeds the contractually defined ranges, the cleanroom fails to function properly (= downtime) and the activities performed in the cleanroom are compromised.

Downtime must be avoided at all costs, as it can lead to interruptions in production activities, with major financial consequences or certain risks for employees working in a cleanroom as result. In order to avoid downtime, the performance of the entire cleanroom installation (even its individual components) must be closely monitored and failures must be detected and resolved directly.

Affiliated to ABN Cleanroom Technology is the company Smartlog. Since 2013, this company is active in the field of Industry 4.0 and specializes in HVAC asset diagnoses and laboratory & pharmaceutical applications. By connecting sensors, the Internet of Things and a cloud-based platform, Smartlog helps ABN transform and connect unconnected business infrastructures. Together, we are able to continuously collect cleanroom data regarding the efficiency, performance and usage of cleanrooms in order to act immediately when something treatens to happen. Being already in a unique position with our CleanConnect IIoT system, we experienced that there is still room for improvement.


Intelligent future thinking machines

Throughout the years, we received a clear signal from current and potential new customers that no one in the market (neither ABN nor competitors) can currently formulate a completely acceptable response to the demand for a 99,9% uptime guarantee without using expensive, fully redundant systems. For that reason, we started a trajectory in which we are developing an intelligent box for cleanroom applications, equipped with a self-learning and predictive-prescriptive model that can maximise uptime up to 99,9% of the total annual operation period.

"We quickly found that we could add tremendous value to our customers if we transformed our data-blowing cleanroom installations into intelligent future thinking machines", says CEO and founder Jo Nelissen. "In critical environments such as cleanrooms, the reliability of production processes is crucial. It is in this field of expertise that artificial cleanroom intelligence will play a dominant role in the near future. AI driven analyses -higher uptime, reduced operating and maintenance costs- will become a game changer in the field of cleanroom technology."

In other words, by making maximum use of sensor data, the uptime or availability of cleanrooms can be maximised by avoiding unplanned downtime and reducing planned maintenance to a healthy minimum by scheduling it at the most cost-effective manner. Only so, one can evolve from condition-based to predictive-based maintenance for cleanrooms.

More articles


Cell and gene therapies research brought to a next level by a scalable cleanroom-as-a-service approach

Having access to suitable lab-space for cell and gene therapies research & development can sometimes be a hassle and cause substantial thresholds in the progress of research. Modular cleanroom technology and digitalisation are making it possible to minimize these thresholds and enable scientists to focus on what really matters, their research. With ABN Cleanroom Technology, a leading supplier of modular and pre-engineered cleanroom solutions, we are helping multiple biotech companies to kickstart and scale-up their research & development projects with our unique cleanroom-as-a-service offering.

The role of smart notifications in a modern cleanroom environment

Nowadays, we are overwhelmed by notifications, especially on our smartphones, where we might receive several alerts every hour. Alerts are notifications about noteworthy situations in your system. They make sure you are forewarned if conditions take a turn for the worse, helping to prevent downtime and disaster to your cleanroom manufacturing process.

Dry rooms for humidity control in Lithium-Ion Battery processing areas

With the massive rise of hybrid and electric vehicles, the demand for new and better power sources is also increasing. Lithium-Ion Batteries (LIB) are currently the most efficient and powerful solution for this. The production of lithium-ion batteries takes place in ultra-low humidity dry rooms and requires a great deal of technical knowledge and skills.