Very recently I have joined IRIS, an advanced engineering company that provides in-line monitoring solutions to various industries ranging from pharmaceutical to chemical and from food to agriculture to name a few. We have got our own product portfolio but also offer custom-made solutions to respond to specific requests from the market. One key asset at IRIS is that we design complete solutions that include hardware (the monitoring equipment) and software with build-in models to analyse the data.

A couple of years ago the company made a step ahead moving towards the Industry 4.0 paradigm, i.e., monitoring processes at a larger scale (a whole plant for instance), therefore acquiring much more data that is modelled in such a way that decision taking can be automatized. We have engaged in a series of projects in which we deal with huge amount of data to enhancedindustrial design, processing or management:

• In PREVIEW advanced artificial intelligence and machine learning techniques support cyber physical production systems developed for plastic injection manufacturing processes monitoring, control and optimization.

• In OPENMIND data is monitored and analysed in order to reduce development times by predicting appropriate parameter sets for the manufacturing of entirely customised minimally invasive medical devices.

• In ProPAT a global control platform is developed for data acquisition, processing and mining in order to measure properties of process streams and products, accurately and in real-time via process analytical technologies.

• IbD will be a holistic platform that facilitates process intensification based on the use of robust data about a process and statistical, analytical and risk management methodologies.

• SHAREBOX will be a platform that uses data mining tools to help companies to effectively and confidently share resources (plant, energy, water, residues, and recycled materials) with other companies in a symbiotic eco-system.

The processing of this huge amount of information require loads of computing power that are usually provided by cloud-computing centres. In order to perform the tremendous number of operations they are required to these centres use a massive resource of energy. According to the NRDC, U.S. data centres consumed an estimated 91 billion kilowatt-hours of electricity in 2013 – enough electricity to power all the households in New York City twice over – and are on-track to reach 140 billion kilowatt-hours by 2020. It could be fine if it was not that most of this energy is dissipated in form of useless heat. Actually, efficient cooling is a major concern in the computing centres and 50% of the electric consumption is used for this purpose. And for a company like IRIS, which by the way stands for the Catalan Industrial and Sustainable Research and Innovation, such high level of wasted energy is also a concern. So let’s see what can be done.

Many projects have flourished in the last years that aims at extracting the heat produced in the data and computing centres to redirect it to the nearby offices and houses, swimming pools or even research arboretum. But all these attractive solutionscome with some losses on the way. Recently I came across an even more interesting idea: a high-performance computing centre that distributes the operations to microprocessors embedded in electric heaters in houses and businesses. By doing so, the radiated heat is directly used where it is needed without the need to transport it from one place to another. Also, there is no need to build a huge data centre and to cool its premises. According to Qarnot Computing, the company that had this brilliant idea, such system allows a 78% reduction of the carbon footprint and brings free heat where installed since the electricity bill is reimbursed to the hosts.

As you can see, there exist many green solutions when it comes to high-performance computing. So whatever your needs are choose your cloud computing provider wisely. IRIS will certainly do.

Picture: René Magritte’s “The treachery of images”

By PhD Erwan Guillotel, PMO Manager at IRIS and IbD Project Coordinator. Originally published in http://www.iris-eng.com/this-is-not-an-electric-heater/