Is Teaching Away the Patent Practitioner's Golden Hammer?

Patent practitioners have a multitude of tools in their belt to deal with obviousness rejections. The trick is picking the right tool for the right set of facts. Teaching away is one tool that can be powerful for the right set of facts. It is one of the clear ways to attack even an amorphous obviousness rejection relying on common sense since it was explicitly blessed by the Supreme Court in KSR. Specifically, as discussed in a previous post, the Supreme Court explained that when the prior art teaches away from combining certain known elements, discovery of a successful means of combining them is more likely to be nonobvious.

But at the same time, our human brains are susceptible to cognitive bias, particularly the “Law of the Instrument”. As Wikipedia explains:

The law of the instrument, law of the hammer, Maslow's hammer (or gavel), or golden hammer is a cognitive bias that involves an over-reliance on a familiar tool. As Abraham Maslow said in 1966, "I suppose it is tempting, if the only tool you have is a hammer, to treat everything as if it were a nail.

When reviewing PTAB decisions, one gets the feeling that teaching away tends to be the patent practitioner’s golden hammer. However, the USPTO takes a strict view of teaching away - the prior art reference needs to do more than indicate a preference for something, or tout an advantage of a different feature than that claimed, for example. On the other hand, if the prior art explicitly states that a certain feature should not be used, that is likely a teaching away for adding the feature. Also, a prior art reference can be said to teach away when it suggests that the developments flowing from its disclosures are unlikely to produce the objective of the claimed invention (see In re Gurley, 27 F.3d 551, 553 (Fed. Cir. 1994)). The problem is that one rarely has such a set of facts in the eyes of the USPTO.

Thus, it is important to recognize the limitations of a teaching away argument. It is not sufficient where the prior art merely lists alternatives - that is not a teaching away from any of the alternatives because such disclosure does not criticize, discredit, or otherwise discourage them. See MPEP § 2143.03(VI), quoting In re Fulton, 391 F.3d 1195, 1201 (Fed. Cir. 2004). Likewise, indicating that a certain embodiment is not a preferred embodiment is not a teaching away.

Reading PTAB decisions lends one to believe that practitioners tend to over-argue teaching away - there are many PTAB decisions where the board dismisses teaching away arguments because the prior art does not actually teach away in the strict sense required by precedent. Rather than presenting a strong argument, mis-using teaching away only makes it easy for the examiner or PTAB to ignore your argument and affirm/maintain the rejection. Often times, there are facts to fashion another argument against obviousness, but that is lost because it is improperly forced into a teaching away position.

As an example, consider Audi’s application 16/077,002 related to a fuel cell stack. Claim 1 on appeal is listed below:

1. A bipolar plate for a fuel cell stack, comprising:
an anode half-plate on which an anode flow field is formed, and a cathode half-plate on which a cathode flow field is formed; and
a first elastic seal and a second elastic seal;
wherein the anode half-plate includes an anode sealing section of the anode half-plate that surrounds the anode flow field;
wherein the cathode half-plate includes a cathode sealing section of the cathode half-plate that surrounds the cathode flow field;
wherein the anode sealing section of the anode half-plate is in direct contact with the first elastic seal and the cathode sealing section of the cathode half-plate is in direct contact with the second elastic seal; andwherein the anode sealing section of the anode half-plate includes one or more sealing projections having a first height greater than the rest of the anode half-plate of the bipolar plate, and the cathode sealing section of the cathode half-plate is formed without a sealing projection and/or is substantially flat and has a second height greater than the rest of the cathode half-plate of the bipolar plate.

One of Audi’s arguments was the the prior art (Farrington) taught away from the anode sealing projection having a first height greater than the rest of the anode half-plate and the cathode sealing section having a second height greater than the rest of the cathode half-plate. This seems like a reasonable argument since the prior art taught that the anode and cathode plates each have peripheral projections and that it may be advantageous to limit the maximum height of the projections to that of the other features on the plates. In other words, limiting the height is directly opposite of the claimed feature (i.e., having the projection with a height that is NOT limited to the level of the rest of the half-plate). But alas, this is not a nail for the teaching away hammer… as the PTAB explains:

With regard to Farrington’s teaching that it may be advantageous to limit the maximum height of the projections to that of the other features on the plates, the Examiner correctly notes that a reference is not limited to its preferred embodiments, but must be considered for all that it teaches and reasonably suggests to one of ordinary skill in the art. Here, Farrington merely expresses a preference for a maximum height of projections 10, 11 to be the same as other features on the plates. In other words, Farrington recognizes that it may be advantageous to provide such a maximum height to these projections, but also recognizes the possibility that the maximum height of the projections may extend higher than the other features on the plates. Such a preference is not a teaching away. Appellant does not dispute or otherwise address the Examiner’s position that Farrington’s teaching of a preference for maximum projection heights being the same as other plate features suggests maximum projection heights greater than other plate features, albeit as non-preferred embodiments. Therefore, Appellant has not identified reversible error in the Examiner’s finding that it would have been obvious to provide Farrington’s projections with heights greater than other features on the half-plates.

This case illustrates that the teaching away argument can be elusive - even when you think you have it, you don’t. So, be careful not to over-rely on teaching away and reserve it only for those very particular situations when the facts cannot lead to any other conclusion. Here are some of the most common reasons the PTAB fails to find teaching away:

1. The prior merely expresses a general preference for an alternative invention but does not ‘criticize, discredit, or otherwise discourage’ investigation into the invention claimed;

2. The prior art has no actual criticism, but is merely silent - thus the argument that because the prior art does not disclose the claimed feature it teaches away from it is self-defeating;

3. Appellant fails to provide adequate analysis under the case law. Appellant fails to assert—let alone show—one skilled in the art “would be discouraged from following the path set out in the reference, or would be led in a direction divergent from the path that was taken by the applicant.”

4. The prior art’s mere disclosure of more than one alternative, even with respective advantages, does not constitute a teaching away from any of these alternatives.

In electronic control systems, modes mean something

One of the hard things about patent drafting and prosecution is that it can be very difficult to predict where a future dispute with an examiner will arise. There are hundred or possibly even thousand of words used just in the claims, any one or combination of which can be the source of a dispute. Even mundane words can be, and often are, at issue. As we saw in a recent post (here), the dispute was over the term “layer.” Today, the dispute is over the term “mode” (Appeal 2020-002330, Application 15/673,952).

Specifically, this Amazon application relates to coordinating autonomous vehicles. Claim 1 is below:

A method of coordinating operations of a plurality of autonomous ground-based vehicles, comprising:

receiving, by a processor, information related to autonomous strategy modes of one or more of the plurality of autonomous ground-based vehicles, wherein the autonomous strategy modes comprise a plurality of autonomous strategy modes including uncoupled strategy modes, permissive strategy modes, assistive strategy modes, or preventative strategy modes, wherein each of the plurality of autonomous strategy modes comprises a respective set of available actions of the plurality of autonomous ground-based vehicles;

receiving, by the processor, information related to aspects of one or more environments of the plurality of autonomous ground-based vehicles;

receiving, by the processor, information related to topological constraints associated with the one or more environments;

processing, by the processor, the information related to the autonomous strategy modes of the one or more of the plurality of autonomous ground-based vehicles, the information related to the aspects of the one or more environments, and the information related to the topological constraints associated with the one or more environments based at least in part on an operational goal for the plurality of autonomous ground-based vehicles; and

instructing, by the processor, modifications to operations of one or more of the plurality of autonomous ground-based vehicles based at least in part on the information processed based at least in part on the operational goal.

The application specifically addresses issues with an autonomous vehicle management system that receives information from the autonomous vehicles related to strategy modes, actions, and the environments. Per the Abstract, the strategy modes may include an uncoupled strategy mode, a permissive strategy mode, an assistive strategy mode, and a preventative strategy mode. The autonomous vehicle management system may then process the received information based on an operational goal for the system as a whole. The autonomous vehicle management system may then instruct modifications to operations of one or more of the plurality of autonomous vehicles to achieve the operational goal for the system.

Amazon’s approach here was not that any specific action was new, but rather that the prior art did not disclose a plurality of autonomous strategy modes that each comprise a respective set of available actions. By coordinating the modes of the various vehicles with the management system could better achieve a desired goal. While the vehicles were still autonomous, they would use different available actions in the different modes.

To reject the claims, the examiner interpreted “mode” in a way that enabled an easier rejection. Examiners often take this tack, and it usually follows the below, somewhat predictable, approach (of course there are variations):

1. Identify a term in the claims that can be “broadly” construed due to a generic definition

2. Assert that there is no “definition” in the specification

3. Assert that the specification merely provides non-limiting examples

4. Note that it is improper to read limitations from the specification into the claims

5. Assert that any uses of the term that might be inconsistent with the broad definition do not exclude the broader definition

Here, the examiner used a google search for the definition to assert that the term “mode” was being interpreted as:

a way or manner in which something occurs or is experienced, expressed, or done (see Google, search term: ‘define mode’).

From there, the examiner used prior art that discussed the way in which two autonomous vehicles cooperate to complete a common task as equivalent to a mode. In this way, the different actions performed by the autonomous vehicles in the prior art each constituted different modes.

While Amazon argued against the Office’s interpretation, it did not specifically provide its own definition. Nevertheless, the PTAB sua sponte looked to the Microsoft Computer Dictionary, 344 (5th Ed., 2002), which defined modes in the software context as the operational state of a computer or a program. And because Amazon’s specification defined the strategy modes for atonomous vehicle operations that “operated in” a particular mode, the PTAB concluded that a proper interpretation of the claim term, even under the Broadest Reasonable Interpretation, would not encompass the cited prior art.

So, it is important to push back on Examiner's citations to dictionary definitions that are not consistent with the use of the terms in the specification. To the extend there is a battle of the dictionaries, context is key. Also, if the examiner cites a dictionary and the Applicant is silent, not all PTAB panels can be counted on to consider more appropriate definitions.