- dash — offline access to API documentation. Useful for those long-haul flights without wifi …
- Gartner’s Top Trends for 2015 — ubicomp, IoT, 3d printing, pervasive analytics, context, smart machines, cloud computing, software-defined everything, web-scale IT, and security. Still not the year of the Linux desktop.
- Byzantine Fault Tolerance — Wikipedia’s readable introduction to the basic challenge in distributed systems.
- Move Fast, Break Nothing (Zach Holman) — Gartner talks about “web-scale IT”, but I think the processes and tools for putting code into product (devops) are far more transformative than the technology that scales the product delivery.
ENTRIES TAGGED "iot"
How Moore's Law applies to drones — a backchannel meditation on drone limitations.
Extrapolation is great fun — especially over technology, where Moore’s Law has conditioned us to expect exponentially falling costs and fast adoption. Applied to drones, extrapolation might lead us to conclude that they’ll fill the skies soon, delivering anything we want on demand. They are, after all, rapidly getting cheaper and smarter, and drone-related announcements get tons of press.
So, where will the drones stop? A few of us meditated on the limitations of drones last week on news that Facebook plans to use them to provide Internet connections to those who don’t have them, and on DHL’s announcement that it would begin making deliveries by drone to the island of Juist, in the North Sea. An edited excerpt of our exchange follows. Read more…
Once we acknowledge nearly everything is insecure, we can engage in a more nuanced discussion about security.
“Yes, we get it. Cars, boats, buses, and those singing fish plaques are all hackable and have no security. Most conferences these days have a whole track called ‘Junk I found around my house and how I am going to scare you by hacking it.’ That stuff is always going to be hackable whetherornotyouarethecalvalry.org.
“Yes, there is Junk in your garage, and you can hack it, and if
you find someone else who happens to have that exact same Junk, you can probably hack that, too, but maybe not, because testing is hard.
“Cars are the pinnacle of junk hacking, because they are meant to be in your garage. Obviously there is no security on car computers. Nor (and I hate to break the suspense) *will there ever be*. Yes, you can connect a device to my midlife crisis car and update the CPU of the battery itself with malware, which can in theory explode my whole car on the way to BJJ. I personally hope you don’t. But I know it’s possible the same way I know it’s possible to secretly rewire my toaster oven to overcook my toast every time even when I put it on the lowest setting, driving me slowly but surely insane.
“So in any case, enough with the Junk Hacking, and enough with being amazed when people hack their junk.”
Everyone is racing to build the topmost layer for home automation.
Everyone’s racing to build the “god platform” for the Internet of Things: the highest, most generalized layer of intelligence and user interface that ties together connected devices and web services.
It’s tempting to look for analogy in mobile phone platforms, where Apple was initially dominant and now enjoys an extremely lucrative and influential minority position against Android. There are some crucial differences, though. For starters, adoption won’t be quite as easy; domestic appliances last for a long time, and nothing consumers have seen yet makes connected laundry seem appealing enough to justify early replacement of a washing machine. And even in cases where replacement is relatively easy, the grandest promises entail stitching everything into a seamless system — replacing just the easy stuff can seem pretty lame. Read more…
Buildings are ready to be smart — we just need to collect and monitor the data.
Buildings, like people, can benefit from lessons built up over time. Just as Amazon.com recommends books based on purchasing patterns or doctors recommend behavior change based on what they’ve learned by tracking thousands of people, a service such as Clockworks from KGS Buildings can figure out that a boiler is about to fail based on patterns built up through decades of data.
I had the chance to be enlightened about intelligent buildings through a conversation with Nicholas Gayeski, cofounder of KGS Buildings, and Mark Pacelle, an engineer with experience in building controls who has written for O’Reilly about the Internet of Things. Read more…
Range, power consumption, scalability, and bandwidth dominate technology decisions.
Three types of networking topologies are utilized in the Internet-of-Things: point-to-point, star, and mesh networking. To provide a way to explore the attributes and capabilities of each of these topologies, we defined a hypothetical (but realistic) application in the building monitoring and energy management space and methodically defined its networking requirements.
Let’s pull it all together to make a network selection for our building monitoring application. As described previously, the application will monitor, analyze, and optimize energy usage throughout the user’s properties. To accomplish this, monitoring and control points need to be deployed throughout each building, including occupancy and temperature sensors. Sensor data will be aggregated back to a central building automation panel located in each building. A continuous collection of data will provide a higher resolution of temperature and occupancy information, thus rendering better insight into HVAC performance and building utilization patterns. Comparison of energy utilization throughout the portfolio of properties allows lower performing buildings to be flagged.