Limitless energy for coin-sized sensors prototyped by MetaVEH project

MetaVEH was endorsed – with funding of EUR 4 million – by the European Union’s Horizon 2020 programme as part of a FET Proactive “Pillar 1 – Excellent Science” call out, which called for the submission of proactive projects in the area of emerging future technologies with, in particular, a focus on reducing CO₂ emissions.

Indeed, the sensors which were developed were not only free of raw materials that are toxic to the environment, such as lead, or that are difficult to source, such as rare earths, but instead used a very common element in the form of aluminium nitride. In addition, these sensors do not require batteries (also polluting) in order to operate. This is the result of so-called “energy harvesting“, which involves harnessing the energy present in the environment in the form of vibrations by the use of piezoelectric materials, which enable the conversion of mechanical energy into electrical energy. In addition, in order to enhance the performance of the energy harvester devices, a new type of mechanical metamaterials (i.e. materials specifically “engineered” to have certain properties and reactions) was developed and patented by the Politecnico di Milano and Imperial College London. These are able to manipulate the transmission of elastic waves, and so can “capture” the wave that is passing through them, forcing it to concentrate wherever the piezoelectric material is located. This phenomenon is known as “rainbow trapping“, a term that reflects the capture of the elastic wave and the separation of different frequencies that is typical of a rainbow.

Energy autonomy is a prerequisite for sensors that are used for monitoring the integrity of structures and infrastructure (bridges, motorways, etc), and which need to be positioned at points that would be very hard to reach for the purpose of replacing batteries.

Raffaele Ardito, professor at the Department of Civil and Environmental Engineering (DICA) and coordinator of MetaVEH for the Politecnico di Milano, explained the origin and the various stages of the project, the results of which depended on close teamwork not only with the various bodies involved, but also between different departments and laboratories within the university.

I was already working on the issue of energy harvesting, a theme we’ve been involved with since about 2013; to the extent that we’d already taken out a patent with STMicroelectronics in about 2015, to exploit magnetic interaction and improve the performance of harvesters. I presented this research at a conference in 2019, which was also attended by my colleague Andrea Colombi, who approached me because he was interested in my speech. He was then a professor at ETH Zürich but subsequently moved to ZHAW Zürich during the project . He told me that this subject could perhaps be combined with metamaterials, a field I was already looking into at the time. A few months after that conference, the Horizon 2020 call went out for technologies with significant decarbonising potential. I heard about this from my colleague Richard Craster, a professor at Imperial College London, with whom we were collaborating because we had a PhD student in common. The call out seemed to be tailor-made for us, and led to the idea for the broader MetaVEH project, which brought together all the different topics that each of us was working on separately: metamaterials, piezoelectric materials for harvesters, and interaction with magnetic materials.

We worked closely with Imperial College London, which developed numerical methods for studying and improving metamaterials, techniques we then adopted to design these particular metamaterials, which were improved and combined with the piezoelectric material to create a prototype at a macro scale that would result in a better performance from the energy harvester.
We needed to find some lead-free piezoelectric materials, and various colleagues at the Department of Mechanical Engineering, in particular Nora Lecis, helped us a lot with the use of Funtasma (the Functional Sintered Materials) interdepartmental laboratory. This is equipped with binder jetting 3D printers, with which we were able to create prototypes in KNN, or sodium and potassium niobate, a newly developed piezoelectric ceramic material.

By using the previous patent with STMicroelectronics, we then worked on the magnetic interaction, improving it by applying advanced methods of calculation and implementing it in laboratory prototypes. Together with colleagues from Polifab, in particular the group of Riccardo Bertacco and Federico Maspero (both from the Department of Physics – editor’s note), and inspired by ideas put forward in other fields, we envisaged new ways of creating a magnetic interaction that was as efficient as possible by using the process of magnetic shielding. With this, the magnetic nuclei are shielded by other magnets with opposite polarities, so as to create an impulsive interaction; it was one of our objectives to create an interaction that was as impulsive as possible so as to excite all the frequencies of the metamaterial, creating working prototypes even in a context of real vibrations.

There was also the matter of electromechanical conversion: the metamaterial is like a keyboard with many different frequencies, and an electrical signal with a certain frequency is emitted from each key. Amassing all the signals is a very complicated operation because one cannot simply add them all together, as there is an inevitable difference in phase even between nominally identical resonators. You need to use a method called rectification which enables the signals to be coupled together so that all the voltages have the same sign. We tried mechanical rectifiers, but were not satisfied with that. So we worked with our colleagues at ETH to develop special electronic rectifiers that we called EMetaNode: a system with a “node” made from a metamaterial, with purpose-built electronic circuits that were able to add all the signals together. A study into this will be published in the March 2026 issue of the highly respected Journal of Sound and Vibration. Finally, in collaboration with STMicroelectronics, we worked on the creation of a prototype sensor on the scale of Micro-Electro-Mechanical Systems (MEMS). What we designed is perhaps the first ever energy harvester with aluminium nitride developed on a micro scale.

Of course, we have emphasised the possibility of using low-polluting materials. Generally, such sensors would involve the use of the piezoelectric material PZT (lead zirconate titanate – author’s note), however this contains lead and is therefore very toxic. Instead, one of our objectives was to look into the use of lead-free piezoelectric materials, and I must say that this was a great success. With the support of STMicroelectronics, we were able to make prototypes using aluminium nitride, a very promising material, and then we began to research KNN in our labs, another green material, and demonstrated proof of concept on the macro scale. Ideas are now appearing in the literature that are also on a micro scale, so we’ve been pioneers in the use of this material, which I believe will be used more and more.

Monitoring structures and infrastructure networks has always been our chief focus, and this seemed to us to be the most promising practical application. Some years back, I’d already done work on pilot projects to install sensors on telecommunication towers: accelerometers, inclinometers, and anemometers… One of the problems was the power supply but, in that instance, there was no lack of electricity. On the other hand, if the infrastructure is in a remote setting , there is no access to the power grid, so you have to use batteries, which are notoriously polluting. So we envisaged this device as a killer application. Our colleagues from Multiwave Technologies, the other company that took part in the project, worked on the data transmission, to discover how the sensor node powered by the harvester could collect data – for example from an accelerometer – and transfer it using wireless technologies with very low energy consumption.

In the meantime, we need to discover the correct way of using the data that is collected in this manner, and that’s another strand in our research. I find myself talking increasingly often with managers, even top managers, about infrastructure networks: people who don’t know how to interpret the huge amount of data they are able to collect. In particular, I’m working with my colleagues Alberto Corigliano and Stefano Mariani (both from DICA – editor’s note) on the application of AI algorithms to interpret these data, a task which is beyond the standard user. A very good option is provided by AI agents, using methods that we’ve been looking at for years, such as reinforcement learning. In some cases, there is such an abundance of data that the “digital twins” model might be successfully applied, to help monitor changes to structures over time and to recommend maintenance operations in response to any damage detected by the algorithm.

Then the MetaVEH research continues, including with regard to devices. The component produced on a MEMS scale using aluminium nitride should be linked with EMetaNode, i.e. electronic management, and then integrated into a sensor and combined with data transmission to create a fully functional system. All these pieces have now been completely integrated on the macroscopic scale and need to be converted to the microscopic one. The aim is to achieve a higher level of technological maturity, using transition projects that attract further financing, in order to get to the market.

In this way, the sensors are much easier to use, more versatile and above all much cheaper. They would also allow you to cover very large infrastructure networks with negligible expense. The sensors themselves are very inexpensive, so the greatest cost is in installing the systems; but then, thanks to our solution, they become practically autonomous from an energy standpoint, without the need for further interventions.