My particular area of research has nothing to do directly with Moore’s Law and advanced chip technologies. We currently use 180-nanometer technology for our health sensors. So, at first sight, you could say that in technology terms, we are lagging 10 to 15 years behind 😉
There are various reasons why we work with ‘older’ technologies. One of these is that the analog models have already been fully developed and now really should be “mature”. The focus for new chip technologies typically is placed on digital and RF, while the analog models and building blocks are only developed at a later stage. And mature analog processes are very important for us. Because all of our measurements (temperature, heartbeat, etc.) are analog parameters. We need technologies with very good analog performances (and hence which provide good signals). They also need to be inexpensive and with ideal features. 180nm is currently the outstanding technology for wearable sensors on account of its good mix of reliability, cost and performance.
But wearable sensors certainly also benefit from more advanced technologies. Using 180nm technology we are able to produce chips of almost 30mm²; but if we want to move towards smaller systems with greater functionality, we need more advanced technologies. Which is why at the moment we are looking at scalable analog circuit architectures in 40nm technology. Smaller chips have a price benefit and also increase the range of applicability. Just imagine, for instance, chips measuring 5mm² or smaller that we can use to produce disposable sensor plasters, or integrate them just about anywhere in the world around us.
But there is yet another reason why we need more advanced chip technology. Health sensors generate a great deal of data that enables us to draw usable conclusions. And advanced technology is essential for this type of data processing. Only a limited amount of processing goes on in the sensors themselves and this processing is only designed to extract the relevant data from the measurements and pass it on. The second level where the signals are processed is in a central device that you always have with you (such as your smartphone or smart watch). This device provides the link between the sensors and the cloud.
The health cloud is the third and final level on which data is stored and processed. Significant processing power and speed are required at this level. That’s where the data of millions of people is stored and where specific trends are uncovered, so that certain pathologies can be predicted. For example, your stress measurements can be used as advance warning signs of a heightened risk of burnout. An app on your smartphone will tell you if and when you are in danger – and also give you coaching options to consider, to be able to cope with the stress you are experiencing. Because the next health revolution will not revolve around the sensors that we wear, but the apps and services developed around them.
Two things are certain: wearable sensors will become just as mainstream as smartphones are today and the importance of data will increase enormously. And for that to happen, Moore’s Law with both its ‘older’ technologies and its advanced chip technologies will be of great importance.
About the author
Chris Van Hoof received a PhD in Electrical Engineering from the University of Leuven in collaboration with imec in 1992. At imec, he became successively head of the detector systems group (in 1998), director of the microsystems department (in 2002) and Integrated Systems Department (in 2004), and program director (in 2007). Since 2009 he is department director and program director of HUMAN++ in the smart systems unit at imec in Leuven and the HOLST Centre in Eindhoven. Integrated microsystems research focuses on the application of advanced technology for the creation of miniature components and subsystems, ultra-low power wearable wireless sensor systems, and smart implantable devices. Since 2000 Chris Van Hoof is also a guest professor at the University of Leuven.