Security for MEMS, Sensors

Napa, Calif. — The role of MEMS and sensors is growing as more devices are connected and more intelligence is added into those devices. But that has created its own set of issues involving security and privacy.

“Our strategic landscape is changing,” observed Cynthia Wright, principal cybersecurity engineer at The MITRE Corp. and CEO of Synthus, during a keynote speech on day one of the MEMS and Sensors Executive Conference. “These MEMS and sensors are everywhere. Security is an ongoing issue.”

There are personal safety issues related to the proliferation of MEMS and sensors in everyday life at home, in the car, and elsewhere, Wright added.

She noted that while these devices  represent “an amazing revolution,” they also  provide “a conduit to our most personal data.”

She noted many of the episodes where cybersecurity breaches have occurred, such as the 2016 cyberattacks carried out with the Mirai botnet. Many of these emerging threats are due to supply chain compromises and their “baked-in vulnerabilities,” she said.

While enterprises, governments, and the public are becoming more cognizant of potential cyberthreats, and legislators are paying attention, “mitigation is easier said than done,” she commented. “Who wants to put a firewall around their toaster?”

The 2018 National Cybersecurity Strategy was issued last month, and the current Congress is considering a number of cybersecurity bills, according to Wright. California enacted a connected devices law this year, calling for manufacturers to correct issues that invite cyber breaches, such as default passwords that users never change.

“Security-by-design—we preach this,” Wright said of MITRE, a not-for-profit organization that operates federally funded research and development centers addressing a variety of challenges to society.

“Design engineers need to think about the whole ecosystem, from chip to cloud, in terms of implementing a system that comprises: An immutable device or non-changeable identity; enabling trusted boot; and ensuring over-the-air updates and authentication can be carried out securely,” she stated. “There are industry guidelines,” she said, such as the Industrial Internet Consortium’s Internet of Things Security Framework and NIST’s Lightweight Cryptographic Project.

The U.S. Food and Drug Administration this month released its Medical Device Cybersecurity Regional Incident Preparedness and Response Playbook, developed in conjunction with MITRE.

CEOs and other senior executives, along with their respective boards of directors, may soon be personally held responsible for cyberattacks that affect customers and consumers. “No one wants to face this,” Wright commented.

“Build it in, don’t bolt it on,” she said of cybersecurity precautions. “You can’t leave it up to the user. It only works if they do it—if it’s too hard, people are not going to do it.”

In wrapping up, Wright called for cyber resiliency at the MEMS and sensor level.

Autonomous and Electric Cars
Jérémie Bouchaud of IHS Markit spoke next on “Autonomous and Electric Cars: What’s in It for Conventional MEMS and Sensors.”

He forecast that automotive MEMS and sensors will enjoy a compound annual growth rate of 7% through 2024. “ADAS is a driving force in the market,” he said. At the same time, “MEMS is losing its grip on the automotive sensor market,” while image sensors and radar sensors make up more of the automotive device market in the near future. Silicon magnetoresistive MEMS, however, will be finding more uses within the vehicles of the future.

“There will be close to 10 million autonomous vehicles in 2030,” Bouchaud predicted. L2 vehicles of the present typically contain five cameras and five radar sensors, and some offer night vision capabilities as an option, he noted.

For the emerging L3 vehicles of the near future, the typical sensor suite will include those cameras and radar sensors, along with LiDAR devices, which may be MEMS-based, Bouchaud commented. L4/L5 vehicles, beyond 2025, will have multiple LiDAR sensors, he added. “The main opportunity we see for MEMS is the LiDAR.”

The analyst said more conventional sensors will be necessary for electric vehicles and hybrids, apart from the autonomous driving considerations. While EVs and hybrids now pack 15 to 30 sensors, electrification and development of next-generation propulsion systems will call for 30 to 60 sensors, Bouchaud added.

Machine learning and sensing for autonomous mobility
With its continuing acquisitions of sensor companies, such as InvenSense, TDK continues to build its “wide portfolio, across many sensors,” said Nicolas Sauvage of TDK-InvenSense during his talk on “Sensing for Autonomous Mobility” at the conference.  “From sensors to full solutions, they all need to work together.”

Why is machine learning becoming critical to many applications? “It’s not just for less coding,” he observed. “Performance and insights improve as they are exposed to more data. We are still very early in the S curve for machine learning.”

The Oculus Quest gaming system for virtual reality has TDK sensors, according to Sauvage. Such applications call for “higher and higher resolution, with less computing,” he noted. In AR/VR/MR, “high-performance motion sensors are needed for fast movement.”

When it comes to outdoor and indoor navigation, “3D navigation allows for higher autonomy value,” Sauvage said.

In the afternoon, the conference turned to talk of “Voice as a New Interface for Autonomous Mobility.” The topics included MEMS-based micro speakers, smart ears, voice interfaces and edge processing, and wearable sensors and circuits for health and fitness monitoring. The MSEC was also scheduled to hear from Greg La Follette, winemaker at Alquimista Cellars, on “Sensor Innovations in Wine-grape Vineyards and Wineries,” a highly appropriate subject in Napa Valley.

The conference resumes Tuesday for a second day of presentations, this time on autonomous mobility applications.