Five minutes to ruin a reputation built over 20 years, as Warren Buffett put it, holds true in personal relationships. On the Internet of Things, reputations can disappear in five seconds. How do we move from merely intelligent Things to a level where devices have to be Trusted?
It’s a very sticky question. I hear from IoT skeptics constantly about fears that devices won’t do something they are supposed to do at the moment they are supposed to do it, or that they will do something they are not supposed to do at any time. Both those concerns indicate we are moving toward a new frontier, well beyond privacy and security – an idea I’m discussing in an upcoming speaking appearance.
This may seem a bit audacious considering that we haven’t adequately solved either privacy or security issues for most IoT applications. Designers are still creating most projects using the intelligent device model, not a trusted device model. Some application segments came to grips with this problem years ago. Industrial designers embraced trust in the form of -critical thinking, employing design and certification practices so computers could be trusted even in dire circumstances. That level of engineered-in trust usually has a steep cost, however.
Historically, consumers had a level of intrinsic distrust for computers. Those barriers have lowered as the PC, smartphone, smart TVs, intelligent car, and more devices are now in many homes. The benefits of intelligent devices normally outweigh the maintenance headaches. If a privacy breach or a security vulnerability is found, it can usually be patched. As long as a fix is provided promptly and explained convincingly, use of the device continues.
Bursting the trust bubble, however, is still a deal-killer for most people. Technologically savvy (mostly) younger users are very quick to decide if a device/service is trustworthy – and they will tell their friends and everyone else on social media what they find. When you ingrain a device in everyday tasks, the ease of use and dependability bar goes way up.
IoT devices are becoming -critical at the personal and business level. How do semiconductor and software teams create that kind of trust without spending a fortune on each design?
For 23 years running, the IEEE sponsors the Electronic Design Process Symposium (EDPS) to air ideas on larger issues in EDA and chip and systems design. I’m honored to have been selected to speak in front of the group this year in Monterey, CA on April 21 with the subject “The Internet of Trust and a New Frontier for Exploration”. I’ll be discussing the scope of the trust problem and some ideas on how to solve it, including expanding the use of FPGA-based prototyping tools. Following my talk will be a panel on FPGA-based prototyping, moderated by Daniel Nenni and featuring Tom De Schutter of Synopsys, Toshio Nakama of S2C, and Frank Schirrmeister of Cadence.
Glancing at this year’s EDPS agenda, I’m not the only IoT technologist there. On Thursday the 21[SUP]st[/SUP] there are talks from Serge Leef of Mentor Graphics, Ken Caviasca of Intel, and a session focused on low power challenges for IoT devices. On Friday the 22[SUP]nd[/SUP], the entire day is devoted to cybersecurity.
This symposium is a unique opportunity to meet and hear from technology teams with less marketing hype, and I’d encourage IoT designers and others involved in semiconductor and embedded software development who can make it to the event to join us there and add your voices to the conversation. There are rates for IEEE members, non-members, and students. More information and registration details are available online:
Drop me a note here on SemiWiki, on Twitter at @L2myowndevices, or on LinkedIn (I’m pretty easy to find) if you’d like to meet me in Monterey. I look forward to seeing you there.