No SEMI Summit would be complete without a lively panel discussion that puts industry experts in the hot seat. The panel at the first SEMI European MEMS Summit, titled IoT Beyond the Hype: What are the Hard Facts?, didn’t disappoint for much as what was discussed, as for what was glossed over; and the ongoing discussion that ensued.
Moderated by Jérémie Bouchaud, HIS, the panelists included:
• Behrooz Abdi, president and CEO, InvenSense
• Peter Himes, VP International Marketing, SITRI
• Andrea Onetti GM Volume MEMS and Analog Devices, ST Microelectronics
• Rachid Asmani, VP R&D, MCE Microphones, Knowles, Corporation
Gartner has placed the Internet of Things at the peak of its Hype Cycle, and many are questioning whether it will get caught in Trough of Disillusionment. One only has to listen to recent concerns and discussions to know it’s at least making its going to at least make a pit stop there.
The panel followed a Q&A format, with questions posed by Bouchaud as well as attendees. As not all the panelists answered all the questions, I’ll focus on some of the highlights:
Q: What is your favorite IOT application that you are working on in your company?
Abdi talked about the struggle for new MEMS start-ups to get to market, and the time it takes to get to commercialization. InvenSense is focused on a process platform based on time software and algorithms that will allow third party sensor companies to fuse sensor data with GPS, WiFi and map matching for the indoor location market.
Himes prefaced his response noting that it’s more than just sensors, but also the solutions developed using the sensors that are compelling, and indicative to the approach being taken in China. “One I like that’s a good example targets the deaf community in Shanghai,to fill a need to interact more regularly with their surroundings,” he said. It’s a sensor-based device that enhances input of the remaining senses to allow the deaf to interact more easily with others.
Q. Are we paying too much attention to wearables? Is the IOT the way this market is going to develop?
Onneti said that wearables are the first test of domain bricks for the IoT. We need to test what can be put together. “If the wearable starts to proliferate in market, we can test the limits of technology,” he noted. Ultimately, he said, it’s just a transition phase. The true IoT goes way beyond wearables. It’s a global technology for local business, but it is based on strong bricks. Sensors are a part of it, but it needs smart intelligence inside.
Abdi predicted that the wearables market will be large but fragmented, and include everything from smart watches and fitness trackers in various forms, to elderly, patient, and child tracking; and head-mounted displays. He said vertical markets are still in a nascent stage. Gaming and virtual reality will be huge. There’s a great opportunity, but we should focus on key common denominators.
Himes commented that wearables are targeted to vertical needs; and Asmani pointed to self-driving cars that have $110K worth of electronics based on sensors interacting with each other, as having enormous potential.
This brought up a question I wanted to ask but didn’t: At the consumer level, (smart homes, cars, wearables) are we really looking at novelty items, affordable to a certain socio-economic class? Much as the designer shoes, handbags, and clothes found in the Prada, Versace, and Louis Vuitton boutiques I wandered through in Milan, it is possible these smart devices may not find their way to the masses the way mobile phones have.
Q. Who accepts responsibility for the accuracy of data in IoT devices? What implication does this have in the IoT world?
Only Abdi answered this question, and he merely speculated that it won’t be much different than what hospitals go through with regard to medical data. The lack of answers indicates to me that no one ever considered this before, or contemplated the accuracy of the sensor data.
And what about tracking devices that are the result of cross-industry collaborations? Jorgen Lundgren, Entegris used the example of integrated RFID tags in Garmin GPS locators, that make it possible to track people. Who is responsible if that data gets into the wrong hands? Unintended circumstances such as this need to be considered, and by all rights, should have already been considered, but they clearly haven’t.
Q. With the recent news of the Jeep Cherokee that had been hacked, disabling breaks and cutting the engine, how concerned are you about cyber security?
Onetti noted that the problem is, the more things that are connected, the more they can be hacked. Calling the IoT “a nice game with lots of fragmented ideas”, there are problems that will “make the game difficult, and become a bet.”
At InvenSense, noted Abdi, they are just starting to touch on security. “It goes both ways; with wearables, is the data coming from the person it’s supposed to be coming from?” He didn’t think that securing wearables would be that much different than from a PC, which has security embedded in the hardware. “There’s lots of standardization we can learn from with embedded devices. IoT security will be even more important than a phone. He said they are aware they have to take on some of the burden for security. “The industry is ignoring it for now, but we have to think about it,” he said.
Himes said its not so much ignoring as knowing what to do with it at the networking level. If everything is encrypted at the sensor IoT level, we’ll be faced with a huge power latency penalty. “That’s a challenge with no clear answer at this point,” he said. “Eventually it will have to be addressed at the infrastructure.”
My concern with figuring it out eventually is, what happens to all the devices and connected cars that are already on the market? Is hacking not an issue for them? Will companies be recalling their products and replacing parts with those with security designed in?
Q. It’s clear the IoT creates business for data analytics and services; GE has generated $1B in services thanks to sensors. But what does it mean for MEMS companies? Who can you retain the value you create?
“We are looking for solutions around use cases,” noted Abdi. “Location is one of those places we can do that. Our strategy is to add more intelligence at the sensor level.” InvenSense’ goal for capturing value is delivering an end-to-end solution with sensors, analytics, and accurate location data.
The approach at ST Microelectronics, noted Onneti, is to diversify its offering and have more sensors to match people’s needs. MEMS companies have to be quick to maximize, he noted.
Q. While the Internet provides ultimate freedom, with the ability to connect anywhere, using any device, it seems the Internet of Things is about gaining control over the user. How do you look at this difference?
Abdi said he looks at the IoT not as a way to gain control, but as ambient computing; allowing users to make predictive decisions and improve productivity in your life; in that way he sees it as providing more freedom. Asmani noted that he looks at in terms convenience and safety, and didn’t think about it in terms of freedom. Himes perspective is that the control is what makes the IoT useful for individuals by providing information that is useful in our daily lives.
Q. How far way are we from real new business for MEMS companies with IoT? Are we already generating visible revenue?
‘Fragmented’, ‘scattered market opportunity’ and ‘slow-growth’ were the words used by all the panelists. Himes noted that in China, half of the engagements are around targeted verticals. But it’s definitely here. “The number of applications is phenomenal,” noted Abdi. “From toothbrushes to screwdrivers; we’re happy to wait and see how things develop.”
So as promised by the panel title, those are the Hard Facts. While theoretically, opportunities in MEMS driven by the IoT and wearables abound, get ready to enter the Trough of Disillusionment, and hope that next year, we will be past that, and climbing the Slope of Enlightenment. ~ F.v.T