IoT & Embedded Technology Blog

5G Edge Infrastructure Market Growing Rapidly Ahead of Standardization

Might Air Flow through Northrop Grumman’s Air-Flow-Through Technology Patent?

A couple of weeks ago, an interesting press release appeared in my Inbox, announcing that Curtiss-Wright Controls Defense Solutions (CWCDS) was introducing a new version of their VPX6-490 DSP board, which utilizes dual NVIDIA GPGPU MXM modules and NVIDIA’s Fermi architecture. Besides the fact that this board appears to have extraordinary capabilities, this new version also utilizes Northrop Grumman’s air-flow-through (AFT) cooling technology, which CWCDS was the first to license. This is also, in my opinion, a valid and exciting technological development, providing superior cooling while protecting sensitive components from potentially damaging contaminants in the ambient air. The board release announcement was dated June 28, 2012; the press release announcing the license arrangement was dated April 19, 2012.

Being curious, I decided to poke around a bit and learn more about this AFT technology. I downloaded Northrop Grumman’s patent no. 7,995,436 (click on “Images” to see the drawings), and stumbled on something that gave me pause.

Bear in mind that I am not a patent professional. However, in my career I have been sole or co-inventor on 17 US patents, and thus have been involved in the process and have seen some of its pitfalls.

A Hole in the Patent?

In the Disclosure of the Invention section of the patent, which appears in Column 2 of page 15, the patent reads “…at least one removable module…” (for example, board) “…electrically or electro-optically coupled with the base unit…” (for example, backplane). No problem here. However, in the Abstract, which appears on page 1, and in the claims which appear on pages 18 and 19, the “electrically coupled” option has been omitted. All of the independent claims (Claims 1, 8 and 16) stipulate that the boards (I’m using my own “layman’s” terms here to avoid getting bogged down in the legalese) are “electro-optically coupled” to the backplane. To my knowledge (and I checked into this to be sure) electro-optical coupling is accomplished through the use of light sources (today, LEDs) and photoreceptors or photocells of some sort. Fiber optics, as in VITA 66, could also provide the proper coupling. Any of these would necessitate use of a special backplane rather than the “conventional” VPX types.

Further perusal of the Disclosure of the Invention failed to find any additional mention of the coupling means, be they electrical or electro-optical. I can’t see any reason why AFT would not work with electrically coupled boards and backplanes as well as with those that are electro-optically coupled.

I can envision two scenarios which may explain the omission of the “electrically coupled” option in the patent claims. The first is that this was a mistake, that the electrically coupled option had been dropped inadvertently. The second is that it was deliberately omitted in a response to a PTO Office Action, probably a “103 rejection” (where the invention is deemed by the Examiner to be “obvious to one with ordinary skill in the art”). In a response to this type of Office Action (which is very common), patent attorneys often limit the claims in order to further differentiate the invention from the prior art.

Possible Impact on the Participants

Now the big question is just how this will impact Curtiss-Wright and Northrop Grumman. As I stated above, I’m not a patent professional; the following are my opinions only, and the situation should be sorted out by the attorneys of the parties involved.

I should state here that this issue has no negative impact whatsoever upon the technical merits of either CWCDS’ product or Northrop Grumman’s AFT technology. Any impact would be legal and, perhaps, financial.

In either of the scenarios that I outlined above, the use of AFT technology in CDCDS’ boards is not prohibited even if electrical, rather than electro-optical, coupling is used. However, electrically coupled boards are, in my opinion, not protected under the patent. Curtiss-Wright has gained technical support and relevant expertise from Northrop Grumman through the license, but has not gained any real degree of protection for the cooling technology except in the case of electro-optically coupled devices. I’m not certain whether a court would uphold royalty payments to Northrop Grumman under the license if AFT were to be used on electrically coupled board and backplane combinations. Again, that’s for the lawyers to sort out.

In the first instance mentioned above, wherein the omission was inadvertent, I do not believe that CWCDS would have any unexpected financial exposure. However, under the second instance, where the omission was deliberate, the story may be a bit more complicated.

I have not looked at any of the prior art at all. Forty patents are listed as cited references; twenty-three of these were cited by the Examiner (the cited patents have some relevance to the patentability of the invention; the others listed are for the purpose of providing background). It is possible that one of the cited patents dominates the Northrop Grumman patent, meaning that it discloses a cooling method substantially similar to AFT. In that case the electrically coupled option may have been intentionally dropped from the claims in order to provide differentiation, and thereby make the AFT invention patentable. It is then possible that CWCDS, if they use AFT on electrically coupled boards, may be bordering on infringing this earlier patent, and thus it is conceivable that they may have some financial liability. I would suspect that, in this case, CWCDS would be held harmless and the liability assumed by Northrop Grumman because of their failure to adequately disclose that their patent didn’t cover electrically coupled devices and that another did, but one never knows. In any case it appears to me that it’s time for the lawyers. I’ll be interested to see how this plays out.

Back to Top