Sensors are the heart of the IoT — and printed organic sensors can be used in ways that others cannot. They are lightweight, flexible, stretchable, and soft, so they can bend, twist, or conform to any surface. They can be laminated onto fabric or provide a soft interaction between a robot and an object. New applications are evolving as these sensors become more sophisticated, reliable, and inexpensive. They can be productively used in areas such as communication, information processing, security, medicine, biomedical research, and environmental health. They are more sustainable since they are chemically synthesized rather than using materials that are mined from the earth. That means they can be used to make biodegradable or recyclable devices. Three of the more common types of printed organic sensors are for pressure, temperature, or gas.
Roll to Roll Sensor Printing
A relatively young company, started in 2008, InnovationLab (Heidelberg, Germany), has developed a roll to roll system to print a wide variety of sensor types using specially designed inks. The sensors can be synthesized for a particular application, using piezoresistive, piezoelectric, or capacitive technologies. Key to this approach, is that it can be used to prototype small quantities and once the design has been finalized the process can be immediately transferred to industrial scale and mass produced, both inexpensively and at high speed. That is accomplished by using standard roll to roll label-printing machines modified to implement known printing techniques such as inkjet or screen printing with functional rather than graphical inks.
Printed Sensor Arrays
InnovationLab prints the sensors in the form of a matrix onto a flexible film, typically PET, PEN, or TPU — with the thickness of the film on the order of microns. The sensors are arranged in a matrix of up to a million per square meter. The matrix of sensors functions something like pixels in an image, so pressure sensors, for example, can sense not just the presence of pressure, but it can also dynamically sense its amplitude and location.
The array is formed by printing silver stripes on the plastic substrate. First, a layer of horizontal stripes (rows), then an array of sensors, and finally a layer of vertical stripes (columns). Each printed sensor is at the junction of a row and a column and can therefore be identified as a node at a unique location of a matrix. For pressure sensing, the sensors are force sensitive resistors (FSR), which change their electrical conductivity as a function of the applied pressure. You can therefore use that property to deduce the spatial distribution of the forces on the foil.
You can use that, for example, to identify different objects based on their pressure patterns. Dr. Florian Ullrich, business developer at InnovationLab, demonstrated that by observing the pattern, you can tell whether a bottle is standing or lying. The data of just how much pressure is at which exact location can even be used to identify different objects based upon their pressure “footprints.”
Dr. Alexey Sizov, Head of System Integration and Product Development at InnovationLab, explained that “If the rows of our matrix are driven, we can read the voltage levels at the columns. We use a multiplexer to switch the rows and columns so that each pixel is read in the range of microseconds and sequentially output to a high-speed ADC.”
The data outputs could be sent via CAN bus, or USB, or wirelessly, to a computer for visualization and analytics and also via a gateway to a cloud.
The data from the InnovationLab printed sensor array can be sent by CAN bus, USB, or wirelessly directly to visualization and analytics software residing in a computer or via a gateway to be networked with other IoT devices and/or sent for analysis to a cloud. For each particular application, you decide which data you want to send. The outputs of a number of sensors can also be connected in series to a single electronics processor. That may not be as powerful as utilizing individual electronics for each sensor matrix. On the other hand, if you are designing for a specific application, it could simplify things.
There is a great range of possible applications, for example, pressure sensors can be “tuned” to respond to a wide range of forces, from a few grams to a couple of hundred kilograms. And they can be printed at densities of up to a million per square meter.
Smart beds. One application that could not practically be addressed with conventional sensors, is a smart bed. First of all, the printed sensor array could have a large enough area to cover the surface of a bed. The sensors are so thin, that a person sleeping on them would not know they are there. It could be printed on TPU, thermal plastic polyurethane, which has good elasticity, is transparent, and resistant to oil, sweat, grease, and abrasion, and has a nice feel to it. You could even buy it with a thermal transfer foil so you could iron or laminate it onto a bedsheet. An important use for this bed monitoring would be locating pressure ulcers (bedsores) for a hospital patient, which are a major source of serious side effects during hospital stays. These are formed when a patient lies in one position for an extended period of time. But they can be treated, or even prevented, if they are detected early. Printed pressure sensors can be designed to be sensitive enough to detect and localize pressure points on the body and to measure their time duration. These are the factors that would indicate in advance whether an ulcer is starting to develop. In a hospital or nursing home, the data from a number of beds could be networked, transmitted to a server for analysis, and then pushed to nurses' smart phones or to a monitor at a nurse's station — in real time — to determine when a patient should be turned. It could also be stored for later analysis and record-keeping. Although these factors are different for each individual, without real-time information all patients would have to be turned at the lower limit, perhaps every half hour. That would be an unnecessary waste of nurses' time and energy.
Smart carpets. Another IoT application arose because of the social distancing requirements for COVID-19. InnovationLab produced a “smart carpet” mat, which they installed in a large supermarket to promote social distancing and control the number of customers allowed into the store at any one time. Because of the high density of the sensor matrix — more than 8000 sensors spaced at 1 cm — and their sensitivity, the data sent to a processor is able to distinguish between a person and a shopping cart. And because it is lightweight and flexible and connects wirelessly, it can be rolled up and carried by hand to different locations.
Smart factory. By lining stocking shelves with sensor arrays, you can track the fill levels of the stock in a factory, see it on a central screen, then order prior to a shortage. With an increased level of integration, you could autonomously generate a refill order. This would pay for itself by keeping production running without downtime.
Further sensor types could detect fill levels, touch events, temperature, moisture, and gases.
Smart warehouse. Sensor foils placed throughout a warehouse could be used to keep track of all of the goods stored there. For example, there are picking places, where goods are exchanged. A person might receive an online order to pick up three items from shelf A and two from shelf B. Weight sensing foil underneath the piles of goods would output a signal confirming that the proper number of items were removed from the correct locations. For this application, the foil outputs would be connected by CAN bus to central processing electronics that sends out the digital data.
Automotive charging stations. If someone driving a hybrid or all-electric vehicle needs an immediate recharge, pressure sensors at each parking spot could send out information about which, if any, places are available.
Battery health monitoring. Battery cells expand and contract during charge/recharge cycles. A sensor coating could detect that and use the information to balance cell use, prevent overcharging, measure temperature, and by these means optimize the battery life.
Car seat monitor. The foil can be integrated into a car seat to measure the force profile of a person and analyze it to sense a person's sitting position. By training the AI, you could even identify which regular driver is sitting in the car and adjust the seat and steering wheel positions appropriately. The information can also provide a basis for various driver assistant and safety systems, e.g. seat belt reminders and emergency call systems. Sensors can detect whether a children's seat is located on the passenger side and will automatically deactivate the airbag.
From R & D to Production
In order to develop, and then produce, printed sensing foils for a particular application, one needs to go through a series of stages. InnovationLab has facilities for developing and then commercializing printed sensor products. They provide R & D services and a roll to roll press to produce pilot runs. For the final production run, they have a partnership with Heidelberger Druckmaschinen AG, the world market leader in the manufacturing of printing presses, whose factory is located nearby.
InnovationLab has a highly modified Gallus RCS 330 printing press that supports prototyping and pilot production of up to one million (finger-sized) sensors per day. The press can accommodate substrate widths up to 33 cm and unlimited length. It can use screen, offset, flexo, or gravure, as well as an option for inkjet, printing processes. Heidelberg's production site features a more highly developed Gallus RCS 430 printing press that is solely used for the industrial production of printed sensors, run in a three-shift operation.
A critical piece of the design process is developing the right ink for each application. For that, InnovationLab partners with major suppliers such as BASF SE.
Why Printed Sensors?
First of all, we've outlined some of the unique applications for the variety of sensors that can be printed on plastic foils. Their key characteristics are their flexibility, lightness, and low cost. By using a roll to roll printing process on a modified standard press, you can go right from a pilot run to mass production with minimal effort and expense. Once the initial costs have been paid off, as the production quantity increases the cost per foil is vastly reduced, mostly determined only by material costs and printing speed.
This article was written by Ed Brown, Editor of Sensor Technology. For more information, visit here .