posted by Bryon Moyer
From where we sit in the electronics world, you can sometimes be lulled into thinking we see all that goes on around us. And then you suddenly get introduced to a whole new area that operates all on its own, with its own expectations, standards, and even language. (“Troffer” is used no less than seven times in the release I’m about to reference. It’s a portmanteau of “trough” and “coffer” – a fixture for fluorescent lights. But I digress.)
The thing is, these once-independent worlds are starting to come together. The one we’re talking today, as you’ve probably guessed, is lighting. Commercial, to be specific. And there are actually two different worlds here: the world of controls, which has largely focused on HVAC and other systems that long ago acquired more than just on/off switches, and the world of lighting.
LEDs have changed lighting forever. They provide opportunities both for savings and for control that prior technologies can’t handle. Combine that with an apparent Title 24 mandate pushing commercial lighting to be both dimmable and accessible by the utilities, and you have now brought controls and lighting together.
Specifically, we’re talking about Daintree, a company with a history in controls, and LG, a company that makes… well, we won’t list them all here, but… amongst their weaponry are such diverse elements as LEDs (and a bunch of other stuff). This not only brings controls into lighting, but also raises the possibility of merging the controls of HVAC and other systems together with lighting for more holistic energy management.
Their recent joint announcement addresses a fundamental challenge with change like this: not only the cost of installing new plant, but, in particular, the cost of retrofitting existing buildings. They claim installation cost savings of 85% and resulting energy efficiency improvements of up to 90%. Nine-oh. That’s… a lot.
And the key to this is simple: wireless. It is possible to install wireless network circuits next to existing light fixtures, but Daintree and LG have gone further than that: LG has integrated the wireless capability into the LED drivers. So a single unit will provide both the power and the networking. Which makes a ton of sense, because ultimately it’s the power that the network is controlling.
You can see, of course, how wireless saves installation money as compared to using a wired network. Not even sure I need to describe what would be needed to run a wired network to all lighting fixtures… With wireless, you install, “pair up” – basically, give all the nodes an identity in the mesh, and you’re off.
They’ve chosen Zigbee for its low cost and low power, saying it was the simplest, most cost-effective solution. Unlike some of the descriptions I’ve heard, they found it “streamlined” – as long as you’re not moving lots of data. Which, of course, they’re not. And it has a meshing capability built-in (we’ll talk about meshing with BlueTooth in a future piece).
These types of solutions will also eventually make their way to residential lighting, although they say that the sale will be different. For commercial, it’s all about ROI. I’m assuming that the ROI story would hold with homes as well, but that may not be the way they appeal to buyers.
You can read more about the joint Daintree-LG solution in their announcement.
posted by Bryon Moyer
Coventor recently released a new version of SEMulator 3D. We’ve looked at this tool before; it’s what they call a virtual fabrication platform – helpful for simulating semiconductor processes.
Featured in this upgrade is an improvement in the modeling of so-called pattern-dependent etch effects. In other words, how an etch proceeds at one spot depends on what’s around it. And looking farther out apparently makes for a more accurate simulation result, so, with this release, they’ve increased the radius that defines the region or neighborhood to be evaluated when assessing what the local layout looks like.
They’ve also sped up their etch simulation in general.
Meanwhile, they’ve more fully productized a couple of existing features. One is a structure search capability. This allows the user to find a specific structure in all of the various models. This can be particularly useful, for example, when you learn about some particular yield-impacting configuration and want to figure out which models it affects.
The other relates to their Expeditor tool, which is effectively a design-of-experiments assistant. Its older incarnation was as a spreadsheet-and-command-line tool. Which is great for hardcore “GUIs are for weenies” users. But, apparently, weenies want to use it too, so the new release features a full-on GUIfied version.
You can find more details in their recent announcement.
posted by Bryon Moyer
In my role, you pay attention to conferences. Ideally, you’d go to them all and sift through the repetition to get interesting golden nuggets to write about. But, especially outside Silicon Valley, that means lots of travel, and you start to pick and choose.
As an actual practicing engineer, you (hopefully) have actual projects to work on, so traveling to conferences means trading off some productivity for learning useful things. Question is, how do you choose?
In particular, the IoT is blowing up huge. It’s the way you get attention or traction these days. And I have to admit, I’m part of the problem:
Diligent PR person: “We have some news about [something] that we’d like to brief you on.”
Me: “Well, that’s actually not my area of coverage. I’m focusing more on IoT and related technologies these days.
Diligent PR person: “Oh, this totally has an IoT play.”
And you wonder why every press release mentions the IoT. OK, guilty as charged.
But when it comes to conferences, well, I can’t even keep track of all of the IoT conferences. Just in the Bay Area alone:
- EE Live had a collocated IoT Summit earlier this month.
- There was a “Markets of One” conferences in Palo Alto (although that’s more for designers, but it’s related) a few days ago.
- There’s the Internet of Things Developer’s conference collocated with the Multicore Developers Conference in May.
- Today I just saw that Internet of Things World is happening in Palo Alto in June
All presumably put on by different organizations. And that’s just within, what, three months? All in the same geographic area? Heck, there are probably others I’m unaware of.
You wonder, “Does the world need that many IoT conferences?” But then again, it appears we have a free market at work here. The world may not need this many, but markets aren’t about what the world needs. Presumably, each of these tries to attract the most attendance with the best presentations and the catchiest hooks and the most effective marketing. Never mind that there are many faces you’ll see presenting at multiple conferences, nullifying some of the differentiation.
I’ll be curious to see how many of these there are next year. I sense a bit of IoT fatigue – not that the IoT will necessarily go the way of push technology, but that we’re getting used to the idea of the IoT, so it’s not going to be a hook for much longer.
But my final question is, with this many conferences on the same topic, how do you pick and choose which ones to go to? Do you line them all up and compare programs? Is it pure convenience on timing? Is it the reputation of the organization putting on the show?
I’d love to hear your thoughts on this in the comments.