x

Umachines Claims 2D MEMS Advance

Component startup Umachines Inc. plans to use the upcoming Optical Fiber Communication Conference and Exhibit (OFC) to showcase its particular take on MEMS (micro-electro-mechanical systems).

Umachines is part of a growing roster of companies lining up in favor of 2D versus 3D MEMS (see Umachines Unveils MEMS Switch and Umachines Is Wild About MEMS). Unsurprisingly, the startup says its take on 2D MEMS sets it apart from the rest.

Putting this claim into perspective requires some definitions: MEMS is a technique wherein arrays of tiny tilted mirrors are used to shunt light in various directions. Two basic types of MEMS arrays are in use today: 2D MEMS, wherein the multiple mirrors are arrayed on a single level, or plane, in an optical component (see All-Optical Switching Tutorial, Part 2); and 3D MEMS, in which the mirrors are arrayed on two or more planes, allowing light to be shaped in a broader range of ways. In general, 2D MEMS is thought to be cheaper and easier to make than 3D, but it supports a smaller switch port count.

Both 2D and 3D MEMS call for the tiny mirrors to hinge up and down at specific angles in order to direct light through a switch. Therein lies much criticism of MEMS technology in general: The wear and tear caused by moving parts leads to quick obsolesence. And getting the mirrors to angle just right requires very complex controls.

Now Umachines says it's got a way to overcome both these negative factors. Instead of moving the mirrors horizontally to direct light, the vendor permanently sets the angle of the mirror in the manufacturing process. In the finished module, the mirrors move vertically into the path of the light. No angles, no hinges.

This setup, Umachines claims, makes its MEMS hardier and easier to manufacture. "We believe that our [technique] has enabled us to manufacture components that have superior performance and reliability," says Cheng Teik Tan, VP of business development at Umachines. He touts a recently signed manufacturing contract with Sanmina-SCI Corp. (Nasdaq: SANM) as evidence of the startup's fabrication viability.

Is Umachines's take really unique? It's tough to tell. Competitors such as OMM Inc. seem indifferent to news of the innovation.

"Let me say this: We've been around for five years, we ship to over sixty customers. I think anybody coming into this market has a long, tough battle ahead," says Conrad Burke, senior VP of sales and marketing at OMM.

Burke says that over the next couple of years, market demand will focus on smaller-port-count switches for edge and metro use, not massive all-optical crossconnects. Customers are eager for modules featuring up to 32x32-wavelengths, he says.

Like Umachines, OMM has decided on 2D MEMS, since 3D technology is still "expensive and unproven," according to Burke. OMM abandoned its 3D MEMS development plans last year (see OMM-inous News), believing the market was farther out than originally hoped.

The roster of companies joining the 2D march is growing, bolstered by financial conditions that force a choice between the technologies. Like OMM and Umachines, DiCon Fiberoptics Inc. sees lots of life in 2D MEMS (see DiCon Intros 11 New Products ).

This isn't to say 3D has entirely lost its luster. Agere Systems (NYSE: AGR) says its primary focus is on 3D MEMS (see Agere Ramps Up 3D MEMS). Movaz Networks Inc. announced yesterday it will demo its own 3D MEMS module at OFC (see Movaz Sees 3D), and Integrated Micromachines Inc. (IMMI), whose founders are friendly with OMM's, plans to unveil its new 80x80-wavelength 3D MEMS switch at the show.

Last August, IMMI opted to get out of 2D MEMS development. CEO Denny K. Miu claims it was a coincidence that his buddies at OMM got out of 3D MEMS just prior to this decision. But he does concede, "We didn't want to compete with OMM and others for low-port business."

Miu says the key to the 3D market is making an optical switch module that can add to, rather than replace, existing OEO switches. "Our customers want to augment what they already have," he says. Relatively small but scaleable optical switches based on 3D MEMS are just the ticket for equipment vendors looking to add high-speed core connectivity to existing multiservice gear, he maintains.

— Mary Jander, Senior Editor, Light Reading
http://www.lightreading.com
trixie 12/4/2012 | 10:47:48 PM
re: Umachines Claims 2D MEMS Advance Mary,

Mirrored arrays are just one specific implementation of MEMs technology. MEMs, as you know, stands for Micro Electro Mechanical, not "a technique wherein arrays of tiny tilted mirrors are used to shunt light in various directions"

Let's not confuse the application with the technology. There are numerous MEMS applications, many of which have nothing to do with mirrors. Automotive air bag accelerometers come instantly to mind. Nanopumps in the biomed world are another.

In the future, there may well be optoelectronic components that use MEMs technology to control light, but will not necessarily use MEMs Mirror arrays to do so....

Just trying to offer helpful, constructive criticism, no disrespect is directed at you, or your article...
dwdm2 12/4/2012 | 10:47:47 PM
re: Umachines Claims 2D MEMS Advance By the way, Ms. Jander, in case you'd like specific references, I'll be happy to send that to you.

Regards,
AR
dwdm2 12/4/2012 | 10:47:47 PM
re: Umachines Claims 2D MEMS Advance Whats in a name?

The name MEMS was formally given to micromachined electric micromotors and mocrorobots in a 1989 workshop titled "Micro-Tele-Operated Robotics Workshop" held in Salt Lake City. In Europe they called it MST (microsystems technology), however, soon MEMS terminology became popular all over.

However, for optical application, scientists at the NRL called it Optical MEMS and they (and others) coined the term MicroOptoElectroMechanical System (MOEMS). Also many people called it just "Microphotonics" or Opto-MEMS.

In the benifit of keeping things less confusing, the LR should keep this distinction clear. Perhaps they should adapt the name MOEMS, because, thats what is being talked about here!

Comments?

Regards,

AR

sridude 12/4/2012 | 10:47:44 PM
re: Umachines Claims 2D MEMS Advance There are a number of optical MEMS components besides switches. Tunable VCSELs, filters and VOAs are already shipping from various vendors.
Tunable filters and VCSELs do use moving mirrors, but the motion is only along the optical axis.
1-D MEMs, I suppose.
- sd
trixie 12/4/2012 | 10:47:43 PM
re: Umachines Claims 2D MEMS Advance That's my point- most people hear MEMs, they immediately think of arrays of thousands of little mirrors.

Just trying to educate the market a bit.
dwdm2 12/4/2012 | 10:47:40 PM
re: Umachines Claims 2D MEMS Advance trixie: Just trying to educate the market a bit.

Or is it educating the readers and reporters?

The industry is feeding the LR (and the market) what they think their terminology should be called. But unless some systematic definition is followed, it is producing a lot of confusion.

My recommendation is the LR should take it a bit seriously.

AR
trixie 12/4/2012 | 10:47:24 PM
re: Umachines Claims 2D MEMS Advance DWDM2: Or is it educating the readers and reporters?


They would be constituents of the market, no?


As with any fledgling industry, standards and conventions become commercial battles, the victors being the ones who get to call the shots via De facto standard adoption (money talks, as always)

I agree with you that some conventions need to be established, as well as some discipline in their use, but I doubt this will happen until we see a market leader (I mean dominant, not just the highest % share in a tiny market).

Of course if other MEMs technologies pop up that don't use mirrors to switch light, the need to differentiate the two will drive this to some degree.

I would also go so far to state that MEMs technologies will become a major element in optoelectronics for this industry in the coming years, as it matures and minaturizes. This will certainly drive the standards.


dwdm2 12/4/2012 | 10:47:21 PM
re: Umachines Claims 2D MEMS Advance trixie: I agree with you that some conventions need to be established, as well as some discipline in their use, but I doubt this will happen until we see a market leader (I mean dominant, not just the highest % share in a tiny market).
-----------------------

No argument here. But the point I was trying to make (especially for the LR!) that: "why try to reinvent the wheel?" The nomenclature is adapted by the pioneers in appropriate forum and proper references are available.

If LR (and the industry for that matter) is not using the terms that are already defined the inferences are (in my assessment):

i) Either the LR/industry is not doing their diligence...

or ii) They are trying to redefine the terminology...

or iii) ... (take your/LR's shot here)

Regards,
AR
dwdm2 12/4/2012 | 10:46:54 PM
re: Umachines Claims 2D MEMS Advance It would be interesting to see if Ms. Mary Jander has anything to add to the MEMS/MOEMS nomenclature issue!

Regards,
AR
HOME
Sign In
SEARCH
CLOSE
MORE
CLOSE