Industrial Internet of Things (IIoT) is often referred to as Industry 4.0 to denote the fourth industrial revolution, that of connecting products, machines, services and humans through the cloud. IIoT is moving toward an era of increased interconnectivity, which is forecasted to soon outgrow consumer IoT. This new age of smart manufacturing requires new ecosystems to facilitate full factory automation and real-time monitoring with the aim of increasing productivity, enabling predictive maintenance and optimising supply chain and asset traceability whilst developing a safer and more secure environment.
Interconnection of machines requires the deployment of sensor nodes, devices that take measurements (for example, the location of a valuable asset, the environmental conditions in which a food product is being stored or the heat emitted by a damaged pump), which can be stored or transmitted to a hub for further processing. A lot of sensors are currently cabled, which is often impractical when pieces of equipment need to be moved around and full re-wiring is then needed. Retrofitting aging production lines with sensors can be tricky, if not totally impossible, when data needs to be acquired on moving parts or near hot machines. For monitoring infrastructure like pipelines or train tracks, the cost of wiring these cables may be too high to make it feasible. So, in many cases, making the device autonomous in terms of the way it is powered, for example by using batteries, makes a lot of sense.
Using batteries also helps with a key requirement of industrial sensors, their size. As the number of these sensing nodes is expected to multiply rapidly, new requirements are put on their size, with a move toward miniaturization so that they become unobtrusive, almost invisible and can be placed in hard-to-reach places, on a turbine blade for example. With less and less space available for the battery, it is difficult to provide enough energy to the device and allow it to take measurements and transmit data.
This is even trickier in the generally hostile industrial environment, where sensors (and the batteries powering them) are expected to work 24/7, near constantly moving or vibrating machinery, in sometimes high moisture and dusty atmospheres. Temperature may also become very high — machines may reach 100°C in the textile industry or 150°C in the plastic packaging industry; or very low, near a pipeline in Alaska, or nearer to home, on a February morning when your car was left parked outside. “Industrial” or “automotive” grade batteries do exist for this purpose, of course, but these are generally more expensive and bulky. That’s because they are packaged and protected robustly because the last thing you want is for the battery to leak, catch fire or even explode.
Vibration sensor on manufacturing line, powered by solar panel and solid state battery. The insert is a close-up of an autonomous sensing device, powered by a solar panel and solid state battery, possibly picking up vibrations to ensure full working order of the machine.
One advantage of batteries is their relatively low cost. But what is important is the total cost of ownership of a device. Of course, coin cells are now heavily commoditized and can be purchased for less than a dollar usually. The real cost goes up significantly, however, when you add in the costs to pay a technician to change a failing battery. In a factory that might have hundreds or even thousands of sensors installed, this could end up being someone’s full-time job. But beyond installation costs, what is even more costly to the company is the period when the battery is about to fail or has failed. By the time this has been noticed and the battery has been changed, key data may have been lost. In reality, most batteries only last two to five years, which is too short compared to the expected life of the device they power, which is perhaps 10 to15 years. So, a way of powering smaller and smaller devices “perpetually,” that can acquire data and send them wirelessly to a data hub, would be ideal.
One solution is to power these sensors by combining a small, but long-life, energy storage component, such as a rechargeable solid state battery, with an energy harvester. There are many ways of harvesting energy from the environment, from solar energy using photovoltaic panels, or heat using thermal electric generators (TEG) and vibration with piezo electric devices. Energy also can be harvested from radio or ultrasound signals. So, if the energy budget is balanced correctly, these two components, the battery and the energy harvester, do not need to be large to work for a long time. For example, the solar cells can provide just enough energy to the sensor and to charge the battery during the day, whilst at night the battery can provide the power.
In fact, now is an exciting time for ultra-low power electronics. On one hand, the efficiency of energy harvesting devices is constantly increasing, with high efficiency indoor solar cells which can provide at least 20 mW/cm2. On the other hand, sensors and power management integrated components are becoming increasingly less power hungry, some are consuming less than ~1 mW. New radio transmission technologies and protocols are also appearing, helping balance package size, range and power consumption. Bluetooth Low Energy for example is often used for short distance low power transmissions, with only 5 to 10 mA current pulses required.
Then, there is the problem of the energy buffer. I have already mentioned that conventional primary or rechargeable batteries (coin cells, cylindrical batteries, Li ion batteries…) can be a good solution, but they cannot be easily miniaturized due the way they are packaged. Most also operate up to 70°C to 80°C, which is limiting when you need to monitor the efficiency of a drilling tip, which reaches several hundred degrees Celsius. It is the liquid or polymer electrolyte that creates these limitations, and it is also the source of their safety issues and limited longevity. Supercapacitors are another useful option because these have high pulse current capabilities, but typically retain less energy per unit volume than batteries. More importantly, though, supercaps, and to a lesser extent conventional batteries, have high leakage current. This means that a proportion of the energy provided by the energy harvester, which is small, is lost in electrochemical reactions within the supercap, which renders it less efficient. And this gets worse as temperature goes up.
Over the last few years, Ilika has optimized the chemistry and synthesis processes for the construction of its range of solid state batteries, named Stereax.
An expanded construction view of the Stereax solid state-state battery.
These solid-state rechargeable batteries measure a maximum 1 cm2, and are less than 1-mm thick. They also won’t explode or catch fire. Because they don’t have liquid components, they yield long lives, can be recharged thousands of times (if one charge-discharge “cycle” is one day, this equates to up to 10 years’ lifetime) and crucially have ultra-low leakage currents, in the nA range. As explained above, these properties make them an ideal energy storage device when “trickle-charged” by small energy harvesters. The latest Stereax solid state battery is the Stereax P180, a 180 mAh battery which can operate between -40°C and +150°C, thanks to the selection of materials from which it is made. This operating range allows their deployment in Industrial IoT applications, but also automotive, aerospace (near the engine or exhaust) and monitoring of infrastructures such as bridges, pipelines and train tracks.
As the number of these sensing devices is expected to multiply rapidly, new demands arise in terms of size, with a move toward miniaturization, and their performance in increasingly hostile conditions, particularly in increasingly lower and higher temperatures. Small size solid state batteries combined with energy harvesters such as small photovoltaic panels or thermal or vibration harvesters, present a unique solution for long life powering of sensor nodes.