By Pete Singer, Editor-in-Chief
SEMI’s European MEMS Summit will be held on 17-18 September 2015 in Milan, Italy. Over the course of the two-day event, more than 20 keynote and invited speakers from the entire supply chain will share their perspectives and latest updates, including participation by European MEMS leaders. In addition, a focused industry exhibition will complement the conferences offering with additional networking opportunities.
In advance of the event, we asked members of the conference steering committee about what’s happening in the world of MEMS. Answers came from:
- Stefan Finkbeiner, CEO Bosch Sensortec
- Benedetto Vigna, Executive Vice President and General Manager, Analog MEMS, and Sensors Group, STMicroelectronics
- Christophe Zinck, Senior Application Engineering Manager, ASE Group
- Eric Mounier, Senior Analyst MEMS, Yole Developpement
- Martina Vogel, Officer of the Director of the Institute, Fraunhofer ENAS
- Yann Guillou, Business Development Manager and MEMS Summit event Manager, SEMI Europe Grenoble Office
Q: What do you see as the big trends and challenges in MEMS and their applications, particularly with regard to the IoT.
“The application of MEMS sensors to the IoT-enabled markets (e.g. wearables, smart home, etc.) will require sensors to shrink further and to work even more power-efficient as in smartphones,” said Dr. Stefan Finkbeiner, CEO Bosch Sensortec. “In particular, the application side of the sensor will demand more attention. The value-add of a sensor must be convincing to become designed into a certain product,” he added.
Finkbeiner said he sees a big market pull for gas sensors such as the Bosch in-door air quality sensor, the BME680. “That trend is visible for the smartphone as well as for the IoT-enabled markets, like for example the Smart Home market,” he said.
Martina Vogel, officer of the director of the institute, Fraunhofer ENAS, said: “We see, that MEMS exist almost everywhere in our daily lives – in our homes, our cars, our workplaces – and yet they go largely unnoticed. Despite this low profile, microsystems have undergone rapid development in the last two decades, evolving from miniaturized single-function systems into increasingly complex integrated systems. From our point of view we call these complex integrated systems, smart integrated systems.
From performance point of view we distinguish between different generations of smart systems. The first and the second generation entered into diverse applications. The first generation of Smart Systems consisted of several packages of components connected on a single substrate, or printed circuit board. These devices are commercially available in medical applications such as hearing aids and pacemakers, as well as in automotive applications such as airbag systems. The best-known example of a second-generation Smart System is the ubiquitous smart phone, which has seen great commercial success.
Smart systems of the third generation are self-sufficient intelligent technical systems or subsystems with advanced functionality, which bring together sensing, actuation and data processing, informatics / communications. Therefore these systems are not only able to sense but to diagnose, describe and manage any given situation. They are highly reliable and their operation is further enhanced by their ability to mutually address, identify and work in consort with each other. Such smart systems will be the hardware basis for the internet of things (IoT).”
From technology point of view, Vogel said such systems “are not limited to silicon–based technologies but integrate polymer-based technologies, printing technologies (e.g. for printed antennas, printed sensors, displays or batteries), different nanotechnologies (e.g spintronic devices, CNT based devices or devices based on embedded nanoparticles) and even embroidering technologies for sensors.”
Benedetto Vigna, Executive Vice President and General Manager, Analog MEMS, and Sensors Group, STMicroelectronics, said: “The next wave of MEMS development is moving toward actuation and, while the ripples from these beautiful little machines have been building slowly for years, they are converging quickly with the Internet of Things (IoT). We are beginning to see new applications such as tiny mirrors that enable people to interact more naturally with technology, smaller, faster autofocus solutions for mobile phones, and new types of printheads for 3D printing — and this is just the beginning.”
Christophe Zinck, senior application engineering manager, ASE Group, said the big trends and challenges from his perspective are “form factor (especially height), co-integration (flexibility to be used in different modules/SiP (in term of packaging of course but also compatibility with different wireless standard), power consumption and, of course, cost.”
Eric Mounier, senior analyst MEMS, Yole Developpement, said: “For us, MEMS is just a technology among others that could answer the IoT’s requirements for sensors. Indeed, type of sensing required for IoT is very broad: Inertial sensing, chemical sensing , pressure sensing, light sensing … any physical event.
Sensor for the internet of things follow several requirements, Mournier says:
- Low power consumption (Due to the integration in wireless battery powered modules)
- Small form factor (Due to the need for small wireless sensors)
- Low cost (As IoT large expansion lies in the availability of low cost sensors)
For now, several sensing solutions exist in different fields (inertial sensors in smartphones for example). But strong challenges still have to be overcome:
- New sensing solutions (such as MEMS chemical sensors, etc.)
- Low cost, highly integrated solutions (via 3D stacking, etc.)
- Standardization; The IoT is the accumulation of thousands of different applications requiring low cost solutions, but with limited volumes. Developing one sensor per application is not possible due to development costs.
“I am pretty confident MEMS will be used for IoT, specially for gas/chemical sensing. MEMS technologies for gas sensors have many advantages compared to other technologies: Up to 50% size reduction and cost reduction, CMOS scalable technology,” Mournier said. “With cost and miniaturization to be a driving force for consumer and industrial Iot applications, it opens the way to new technologies such as MEMS.”
Q: Sensor fusion is an intriguing thought and the ultimate device might have multiple sensors integrated with energy harvesting, a thin film battery, a microprocessor/ASIC, wireless communication capability, etc. How far away from that are we? What are the big challenges? Is it cost? Integration? Packaging? Form factor? What are the leading applications?
ST’s Vigna said “We are already well on the sensor-fusion path that contains multiple sensors integrated with a thin-film battery, a microprocessor/ASIC, and wireless communication capability. The two technical challenges are low-power radio and high-efficiency (energy) harvester.”
Finkbeiner said Bosch Sensortec already provides leading edge sensor fusion SW integrated within a multi-sensor 9-axis device powered by an ARM µController. “This single package device – the BNO055 – is already available and specifically targeting at motion sensing and orientation detection applications in the IoT-enabled markets. Energy harvesting and thin film batteries might still be a bit too far away from being capable of offering enough energy for this particular use case at reasonably small size. But there’s a lot of research in this area. The challenges? Yes, cost/price is always the main driver. Small size is also important. It allows for small form factor products and better placement flexibility.”
Fraunhofer’s Vogel said there is a lot of work carried out with in ECSEL and especially EPoSS. “EPoSS the industry driven Euroean Platform on smart system integration is just working more than 10 years in this field,” she said. “Big challenges are of course packaging and integration from technology point of view. But also issues like big data handling and data security in the internet need to be solved.”
Vogel said market reports concerning IoT predict two trends:
- Printed electronic systems that will enable – low cost sensing. Printing technologies, such as roll-to-roll (R2R) will enable extremely large volumes and low cost. Also expect disposable devices with a short lifespan.
- Sensor “swarms” for inorganic sensing. Devices will have complete integration of sensing, processing RF, energy harvesting, on single small chip ( <1mm2).
ASE’s Zinck said he didn’t see things going that far, “but each sensor fusion is quite specific and current modules are often using custom ASIC, MEMS, etc. The next big challenge is flexibility for co-integration and this will require availability of bare die on the market, otherwise small and efficient SiPs won’t be easily available if you cannot mix best solutions available on the market (in terms of performance and cost, of course).
Zinck said there are also lots of challenges regarding packaging, including compartmental shielding to avoid parasitic between components, antenna on package (especially for wearable), and test.
Q: We’re hearing a lot about wearables and medical applications, but what about applications in the smart home, smart city, smart grid, industrial and, of course, automotive ?
Vigna said: “There are already numerous applications for MEMS in Smart Environments, Smart Driving, and Smart Things and many of ST’s customers are leading that charge by combining elements of ST’s complete portfolio. We’ve got customers using ST MEMS, MCUs, analog and power, and connectivity products in smart thermostats, smart lighting, smart meters, and Smart Driving applications. If you’re not hearing enough about these, it is only because the wearable and medical applications may be sexier.”
Finkbeiner said: The sensors for the other IoT-enabled markets like smart home, smart grid etc. are available or already being developed … what is lacking is the corresponding infrastructure, that means the upper layers for aggregating, collecting and intelligent interpretation of the vast sensor data and bringing them into the cloud. This will for example require standards how to handle sensor data at an higher, more abstract level. But that’s beyond the domain of the MEMS sensor suppliers. At Bosch we have therefore founded Bosch Connected Devices & Solutions, a business unit which develops complete solutions based upon our MEMS sensors.
Vogel said: “Just several years ago Frost and Sullivan pointed out that smart is the new green. The concept of ‘Smart Earth’ is, in fact, the in-depth application of a new generation of network and information technologies. Smart cities arise worldwide. Global concepts for smart production are under development. The Internet of Things – IoT – including smart grid, smart health, smart city, smart buildings, smart home, smart production and smart mobility provides not only big opportunities but is requesting more highly integrated smart systems from the hardware side. The total number of connected devices is expected to grow rapidly. Electronic components and systems are a pervasive key enabling technology, impacting all industrial branches and almost all aspects of life.”
Zinck said: “Wearables and medical are driving SiPs developments as low power and very aggressive from factor, at low cost are mandatory. Smart home, smart city, etc. are using a lot of MEMS and sensors, but the challenges are not exactly the same, some are similar in particular for Smart home (low power, wireless modules, etc.) but there is less pressure on form factors.”
Automotive is a different topic, says Zinck. “The trend we can see is to go smaller for sure, but for the moment it implies move away from leaded packages to leadless, with specific technology developments like wettable flank QFN. Also for automotive two categories have to be clearly distinguished:
- Non-safety applications (like Infotainment): basically similar trend as consumer MEMS, with more and more sensors in the cabin (uphones, pressure, etc.)
- Safety applications: very robust have to be used, but some “intelligent SiPs” are already available like QFN 7×7 TPMS (featuring an accelerometer + ASIC + pressure sensor).
Q: Europe in general is very strong in MEMS for various reasons. Why does it make sense to have the MEMS Summit in Europe?
SEMI’s Yann Guillou said Europe is home to several strong IDMs in MEMS, and most notably home to Bosch and STMicroelectronics. “These MEMS leaders are often identified as the industry’s ‘Titans’. These IDMs have contributed enormously to the European industry, but they have also benefited from a strong value chain in the region: RTOs, equipment and materials companies, foundries, etc. Having such leaders in the region is definitively a differentiating factor for Europe in a MEMS and sensor industry that is facing mounting competition. With the IoT, many new business opportunities may arise and increase the competition. This might shake up the current state of the industry,” he said.
Organizing such event in Europe was pretty straightforward. We took this decision more than 1 year ago and it looks like this decision was right. Today more than 200 people are already registered for this event and we expect to go beyond. I see lot of non-European companies planning to attend, including many US and Asian companies. Interest is strong in Asia for this event. People from Korea, Taiwan and China will be attending. As an example, we will be pleased to receive the visit of a Chinese delegation interested to develop business and technology partnerships with European companies.