413 F.2d 1187
Application of Robert F. WHEELING.
Patent Appeal No. 8097.
United States Court of Customs and Patent Appeals.
September 4, 1969.
D. Carl Richards, Richards, Harris & Hubbard, Dallas, Tex., Richard K. Stevens, Virgil E. Woodcock, Philadelphia, Pa., James H. Littlepage, Washington, D. C., for appellant; O. G. Hayes, New York City, James F. Powers, Jr., New York City, of counsel.
Joseph Schimmel, Washington, D. C., for Commissioner of Patents; Jere W. Sears, Washington, D. C., of counsel.
Before WORLEY, Chief Judge, and RICH, ALMOND and BALDWIN, Judges.
This appeal is from the Patent Office Board of Appeals decision affirming the examiner's rejection of method claims 1-5 in appellant's application1 "as having no novelty other than mental steps." The board reversed the examiner's other rejection of method claims 1-5 under 35 U.S.C. § 103 as obvious in view of Munson and Rubin,2 Draper et al.,3 and Phister et al.4 As to apparatus claims 13-17, the board apparently reversed both of the examiner's rejections, first, under 35 U.S.C. § 103 as obvious in view of the aforementioned references and, second, under 35 U.S.C. §§ 101 and 102 as lacking structural novelty.5
For the present purposes, we accept appellant's characterization of the invention as stated in his brief:
The invention relates to the production of an optimum value of a system output function which is dependent upon a plurality of variables where the effect of changes in the variables upon the function may be determined.
The invention relates to optimization of the operation of a chemical plant, a refinery, or the like where such parameters as temperatures, pressures, feed rates and feed quality can be varied. Optimum operation of a process may be in terms of a maximum economic return for the unit or may be based upon a minimum expenditure of time where time may be important if the process unit and its operation have a bearing on associated units.
The invention minimizes what would otherwise be a laborious process of evaluating all possible combinations of all values of all of the variables. While evaluation of all possible combinations would permit the identification of the best combination of settings on valves controlling flow rates or heat sources supplying energy to a reactor, etc., the labor necessary to make such an evaluation cannot ordinarily be justified and the time is ordinarily not available.
In accordance with the present invention, an optimum is reached quickly and efficiently by changing each of the variables repetitively in a random manner initially and thereafter making the changes less random depending upon past performance and after each change, sensing the results and adopting the new state as a base for the next change only when the new state is better than the previous one.
More particularly the change in state produced in a new set of values in the parameters which are changed at random is sensed. If the new state represents an improvement, the next succeeding random changes are made from the new state. If no improvement results, changes from the first state are again made until a set of random changes results in improvement.
As improved states are reached, the tendency of the changes to be purely random is modified so that past history in the search for the optimum operating point tends to weight the randomness in favor of the most desirable direction toward optimum. However, at all times, a measure of randomness in the changes made in the variables in the search for optimum is maintained.
The solicitor points out, however, that method claims 1-5 require the use of no particular apparatus, and are related to no particular systems. As disclosed, the method may be performed by an analog apparatus, by hand, or by a digital computer and may be employed in such diverse systems as the processing of seismic signals and a catalytic cracking process.
Claim 2 is representative and reads (numbering added by board; paragraphing, by appellant):
2.  In optimizing a scalar output function of a system where a set of input functions to said system which are independently variable are interrelated and coact to produce said scalar output function, the method which comprises
 generating input signals representative of said input functions,
 generating an output signal representative of said scalar output function,
 storing an intermediate function representative of said output signal,
 changing at least one of said input signals by an incremental amount of random character to form a new set of input functions thereby to produce an improved output signal, and thereafter
 changing at least one of said new set of input signals by an incremental amount, the character of which is of a random selection weighted in dependence upon the difference between the stored intermediate function and said improved output signal.
Although the references cited by the examiner have been included in the record, we find it unnecessary to discuss them.
THE FINAL REJECTION
In the Final Rejection, only claims 1 and 5, not 1 through 5, were rejected on the ground ultimately affirmed by the board. Without mention of any statutory section in regard to that ground, the examiner stated:
Claims 1 and 5 are further rejected as a method of mentally solving the optimization problem and manually resetting the controls of some prior art system in accordance with the mental decisions. The only novelty involved in such a method is in the mental steps, which are not patentable. Applicant's argument on mental processes does not apply to claims 1 and 5, as functions, unlike signals, can be mental concepts. It is recognized that claims 1 and 5 do read on the physical process involved, but they are also so broad as to cover a purely mental process.
THE EXAMINER'S ANSWER
In his Answer, the examiner grouped all of the method claims under the heading "B. REJECTION OF CLAIMS 1-5 AS HAVING NO NOVELTY OTHER THAN IN MENTAL STEPS" under which the only section of the statute he referred to was section 101 and stated:
1. Claims 1-5 are method claims. They relate to the analysis of a system and the adjustment of the parameters of that system in such a manner as to cause the system to operate in the best possible manner. Phister has been added as a citation for its illustration of the type of system discussed by Munson and Rubin. * * * Phister shows a system * * * adapted to be controlled to operate in the best possible manner. Phister's system was to be controlled by a computer, but it could also be controlled by hand, with manual adjustment of the parameter controls and mental analysis of the results of the adjustment. Munson and Rubin recognize the possibility of such hand control in their statement * * * that "this random search method could be followed by hand computation or by digital as well as analog means."
2. But applicant, by his method claims, is attempting to acquire a monopoly on a method of doing this analysis and adjustment by hand. Consider the Phister apparatus with the computer removed. Instead of signals al thru an furnishing input data to the computer, have them presented directly to an operator on meter dials, as has long been done in the control rooms of processing plants. Instead of signals Dl thru Dm being sent from the computer to control the process, have them originate in manually operated switches and dials, as is now done in the control rooms of processing plants. Then the operator could sit in the control room, watch the indicating dials, and adjust the control signals in any manner which seemed suitable.
3. The operator, if he wished, could change the control signals in a random manner, as suggested by Munson and Rubin, and note the control signal values thus randomly chosen which improved the performance of the system. Or the operator, if he wished, could change the control signals in the same direction which had been improving the performance of the system in the previous changes, after the manner of Draper et al.
4. Or, having found the parameters necessary for improved performance by the random method of Munson and Rubin, he might decide that he should change the parameters a little further in the same direction, with the possibility of improving the performance a little more. After that he might again try the random method of Munson and Rubin. The latter sequence was random, then historical, then random; it was an obvious and purely mental, change from the method of optimization offered by Munson and Rubin, and it was a purely mental change from the method of process control used since man first began to observe, then control, processes.
5. The applicant contends that a process is not purely mental if it can be carried out completely by instrumentality. In response to this the examiner contends that if a claim can be met by a purely mental operation, it makes no difference that it can also be met by an instrumentality. Applicant is not claiming an instrumentality — he is claiming a method; he should not be allowed a method claim simply because it can be carried out on an instrumentality. By his method claim applicant is, on the face of the claim, attempting to monopolize a method in which his only improvement is a mental step. If he does not intend this, then his claim is incorrectly drafted. If he does intend this, then the fact that it may also read on an instrumentality is irrelevant.
6. Consider the applicability of the above argument to claim 1. Claim 1 calls for generating input functions (al thru an as they appear on meters in the control room), calculating the value of a scalar function (purely a mental function, carried out in the brain of the operator), comparing the value of the scalar function with the value thereof as calculated at the precedent state (purely a mental function, carried out in the brain of the operator), and progressively varying the input function (with control signals Dl thru Dm). The statement that these changes correspond to the weighted resultant of a random vector and a controlled (historical) vector is * * * merely a statement of how the operator thought in deciding how to change the control signal * *
7. Claims 2-5 read in the same manner upon the usual method of manually controlling a process with the exception of the purely mental steps involved. As with claim 1 discussed above in paragraph B (6), the prior art (for example that represented by Phister) meets all of the physical steps. The CCPA (In re Abrams, 38 CCPA 945, 266 USPQ 266, 267-268) has cited with approval the following rule. "If a method claim embodies both positive and physical steps as well as so-called mental steps, yet the alleged novelty or advance over the art resides in one or more of the so-called mental steps, then the claim is considered unpatentable for the same reason that it would be if all the steps were purely mental in character."
8. For example, in method claims 1-5, that change which is random can be distinguished from a non-random change only by considering the mental processes involved in deciding upon that change. Webster's New International Dictionary, unabridged, 2nd ed. defines random as "without definite aim, direction, rule, or method." The mental processes themselves determine whether any rule or method is followed or whether a definite aim or direction is sought. Only through knowing the mental processes involved can one determine whether a given set of mentally directed results were or were not random. If ten coins are flipped, the results will be a random distribution of heads and tails — it could still be random and be ten heads or ten tails or any number between. But if someone wanted five heads and five tails, and he carefully set the coins to five and five, then the results would be non-random — even though five and five is the most probable random result.
9. Additionally, in applicant's method claims, storing a function is a mental step easily carried out with the human memory.
10. Thus claims 1-5 have been finally rejected as unpatentable over previous methods of controlling a process (as represented by Phister et al., by Draper et al. and by Munson and Rubin), the only true novelty in applicant's method being in the mental steps. Although Phister is added of record in this Examiner's Answer, it merely illustrates what Munson and Rubin assume when they write of a process which can be controlled by hand or by digital means.
11. Applicant also contends that he has patentable novelty in the types of changes which are worked on the parameters of the system and, presumably, that even if the steps of deciding what those changes are to be is mental, the step of effecting the changes, being physical, carries its physical attributes over to the types of changes made. The examiner contends that the type of changes made, are no different, except perhaps in sequence, from the types of changes made in any method of controlling the process.
12. In the event the Board considers that the type of change makes the step of effecting that change a step, with its novelty in its physical aspects, then the examiner contends that the claim does not fall within that class of methods intended by Congress to fall under the heading of "process" in 35 U.S.C. 101. It is obvious that Congress did not intend for the Patent Office to deliberately issue patents which could be infringed by a process which the patentee and the courts cannot distinguish from a non-infringing process. This is one reason for refusing patents to processes in which the novelty is in one or more mental steps. In the present invention the only way to determine the type of change intended is by reference to the mental steps. This is shown in paragraph B (8) above. Thus the methods claimed could not fall within the definition of "process" intended by Congress.
THE BOARD'S OPINION
Apparently not fully comprehending the examiner's position, the board was less than explicit, stating:
As we understand the examiner's position, as set forth in the examiner's Answer, all of the claims have been rejected as being unpatentable over the prior art, and method claims 1 to 5, inclusive, have been additionally rejected under 35 U.S.C. § 101 "as having no novelty other than in mental steps."6
The board proceeded to "treat the prior art rejections first." Concerning the obviousness rejection of method claims 1-5, the board held that "[a]s far as we can determine from this record, the only teaching of combining the random and historical methods comes solely from appellant's application" and therefore reversed the examiner's rejection. The board next quoted a portion from the examiner's "rejection of claims 13-17 as lacking structural novelty." Then, the board reversed as to apparatus claims 13-17 since it was "unable to find any disclosure in the Phister et al. patent or any other art relied on by the examiner in his Answer, which responds to the structure set forth * * * [in the claims]." The board concluded that "we will not sustain the prior art rejections based on 35 U.S.C. §§ 102 and 103"7 which the solicitor has characterized as "all rejective rationales based on the cited references."
The board discussed the rejection of method claims 1-5, which it affirmed, as follows:
We agree with the examiner that claims 1 to 5, inclusive, have no novelty other than in mental steps. In our review of these claims we have considered the "Guidelines to Examination of Programs" published in 829 O.G. 1, August 2, 1966. We are in agreement with the analysis of the law as set forth under the heading "Process" and will apply the principles there set forth to the terms of illustrative claim 2. In so doing, we will use the term "utility process" in the sense used in the Guidelines.
A reading of the terminology of this claim will show that the utilitarian functions performed by the apparatus described in order to practice the described utilitarian method functions are not mentioned. It would be impossible to determine from the terms of this claim what the utilitarian process is or how it is carried out or what any process function may be other than the ones that will be referred to below in the analysis of the claim.
Item 1 is a general indication of the field of mathematical discovery made by the appellant. It conveys no idea of what appellant's mathematical process, as such, may be, nor does it indicate any aspects of a utility process. It indicates at most a mathematical field and certain mathematical characteristics to be considered.
Item 2 has the utility recitation of "generating input signals representative of said input functions." It does not state what physical manipulation is involved. The language itself is so broad as to be readable upon the well known mode of functioning of a person individually carrying out the mathematical process; namely, the particular operation being for the operator with a pencil to make a mark on a paper representing the magnitude.
Item 3 is another utility recitation of generating an output signal. It does not tell how. It does not tell the function that the disclosed process may possibly perform. The language is so broad that the person with a piece of paper and pencil can take a ruler, divider, or follow whatever procedure may be necessary in the analysis and make the number of marks on a piece of paper with a pencil that the mathematical procedure may require, each having a particular meaning as required.
Item 4 calls for storing without indicating how this is done. The storage could be by jotting numbers on a piece of paper.
Item 5 recites the utility operation of "changing at least one of said input signals by an incremental amount of random character." It does not tell how. This could be done by thinking of any number and recording it with pencil on paper.
Likewise, item 6 does not set forth any physical utility method, but merely an operation which an individual could perform mentally by using random numbers judiciously chosen to cause the result to go in the wanted direction.
It will be seen from the preceding that, if the claim is construed in the aspect most favorable under 35 U.S.C. § 101, a utility process is recited in such broad terms as to read upon the well known utility process of someone setting down with pencil, paper, ruler, divider to solve a mathematical problem. The claim also indicates that which is set out in the specification as being a very valuable discovery in the mathematical analysis of information. However, the only utility process indicated in the claim is set out in such broad language as to read on the prior art to which reference has been made.
Thus, the sole difference from the prior art utility method is in the recited mathematical discoveries and mathematical procedures which do not come under the statute. The claim thus presents the question of whether it can be rendered patentable because it solves a new and valuable algorithm not within the statute as the result of practicing a utility method which, as recited, is old in the art.
Thus, the analysis present in cases such as In re Abrams, 38 CCPA 945, 188 F.2d 165, 89 USPQ 266, applies. It is essential to determine what portion of a claim defines a utility process, evaluate such recitation as to scope and compare it with the prior art. If the utility process, insofar as claimed, is claimed in such terms as to read upon the prior art (in the case being considered, the well known procedures normally followed by mathematicians), the claim is properly rejected for the reason that it defines no novel statutory subject matter.
* * * * * *
Turning now to the statutory basis for holding that the statements of desired mathematical or other algorithms are not limiting, we call attention to 35 U.S.C. § 112, third paragraph. In pertinent part it states that in a process combination one or more parts thereof may be defined as a:
"* * * step for performing a specified function without the recital of structure, material, or acts in support thereof * * *"
This statute does not state that it is proper to define one or more steps as a step for obtaining a specified result (the carrying out of a non-statutory algorithm) without any indication of the utility function that causes such result to be produced.
In the preceding analysis of the process claim, it was carefully shown that, insofar as process was claimed it was an old process practiced by individuals from ancient times to date. The algorithm and its solution is stated only in terms of a desired end result of the old process.
We will sustain the above rejection of claims 1 to 5, inclusive.
Thereafter, appellant requested clarification as to whether "the reference to 35 U.S.C. § 112 is a new ground of rejection" under Rule 196(b) since section 112 had not been mentioned by the examiner. The board responded, merely saying that
This case has been rejected under the provisions of 35 U.S.C. §§ 101 and 103, and the reference to 35 U.S.C. § 112 is not a new ground of rejection.
Although the solicitor has advised us that there is no need for us to consider the references cited by the examiner, he urges that
this does not mean that all prior art should be disregarded here, since the board elsewhere noted and relied upon "the well-known utility process of someone sitting down with pencil, paper, ruler, divider to solve a mathematical problem" and "the well-known procedures normally followed by mathematicians". Also, by way of clarification, the board said "[t]his case has been rejected under the provisions of 35 U.S.C. §§ 101 and 103, * * *".
In an effort to enunciate a legally comprehendible position from the present record, the solicitor has attempted to "summarize" the examiner's rejection affirmed by the board as follows:
To the extent that claims 1-5 call for steps within the meaning of "process" in 35 U.S.C. § 101, they are at least obvious within the meaning of 35 U.S. C. § 103.
Since 35 U.S.C. § 102 is not in issue in the present case, we do not perceive the relevance of the solicitor's statements that "simple artillery ranging is an excellent example of the claimed method in actual use" and that "[a]nother actual example of appellant's basic method is video tuning of a TV set after small children have played with the adjustment knobs."
In its brief, appellant stated that
[t]he rejection of method claims 1-5 under 35 U.S.C. §§ 101 and 103 was affirmed on the ground that the method claims "have no novelty other than in mental steps"
and directed its argument to show that "claims 1-5 satisfy the statutory requirements of 35 U.S.C. §§ 101 and 103 and are therefore patentable."
In regard to method claims 1-5 presently on appeal here, the examiner has only made reference to two sections of the statute, namely, 35 U.S.C. § 1018 and 35 U.S.C. § 103. However, in our review of the record, we are unable to discern the slightest indication that the examiner relied upon sections 101 and 103 in combination to form a single rejection. To the contrary, the examiner apparently made two separate rejections of claims 1-5, one under 35 U.S.C. § 101 and another under 35 U.S.C. § 103, the latter rejection having been expressly reversed by the board. Accordingly, it appears from the record that section 103 is no longer in issue before us. Perhaps the solicitor's and appellant's confusion, which is apparent from our discussion of their positions, arises from the previously quoted portion from the board's opinion responsive to appellant's request for clarification. But since the board has made no new ground of rejection under Rule 196(b), it merely affirmed the examiner's rejection which, as we have indicated, is apparently based only upon 35 U.S.C. § 101.
We decline to substitute speculation as to the inclusion of any statutory section other than 35 U.S.C. § 101 — whether it be section(s) 103 or 102 or 112 — for the greater certainty which should come from the Patent Office in a more definite statement of the statutory grounds of the rejections. Statutory basis, as well as issues, must be identified and framed with sufficient clarity in the Patent Office so that we can determine with certainty the areas of our authority under 35 U.S.C. § 144. Cf. In re Herrick, 344 F.2d 713, 52 CCPA 1664 (1965). In the present case, the statutory sections relied upon have not been set forth with sufficient clarity to permit a proper determination of the legal issues here. Cf. In re Chilowsky, 229 F.2d 457, 43 CCPA 775 (1956). Moreover, since both the solicitor and appellant have briefed the case on the same apparent misunderstanding9 that the rejection of claims 1-5 was based on sections 101 and 103 in combination, we find the parties' arguments of little help in deciding the difficult legal issues raised in this appeal. Thus, the present record and the parties' briefs and arguments do not present a case for proper judicial determination since it is not clear to us what statutory sections, taken alone or in combination, have in fact been relied upon by the examiner10 and the board.
Accordingly, the case is remanded for clarification and further proceedings consistent with this opinion.
WORLEY, C. J., concurs in the result.
Serial No. 54,503, filed September 7, 1960, for "Optimization with Random and Historical Vectors." The real party in interest appears to be Mobil Oil Corporation, formerly Socony Mobil Oil Company, Inc
"Optimization by Random Search on the Analog Computer," paper presented to the National Simulation Conference, Dallas, Texas, October 25, 1958. The record also includes an abbreviated form of the paper published in IRE Transactions on Electronic Computers, June, 1959, pages 200-203
U. S. Patent 2,628,606, issued February 17, 1953
U. S. Patent 3,079,079, issued February 26, 1963, on an application filed May 21, 1958
In his Answer, the examiner stated, under the heading "REJECTION OF CLAIMS 13-17 AS LACKING STRUCTURAL NOVELTY," that "[t]his is a rejection under 35 U.S.C. §§ 101and 102" (emphasis added).
However, see n. 5 and text associated therewith
Although the board's disposition of the rejection of apparatus claims 13-17 under 35 U.S.C. §§ 101 and 102 is not entirely clear from the record, neither appellant nor the solicitor have discussed the matter. Thus, we will assume that the board's reference to 35 U.S.C. § 102 quoted in the text, supra, encompassed the rejection under 35 U.S.C. §§ 101 and 102. Since the board did not affirm any rejection of claims 13-17, such an assumption is not in conflict with Rule 196(a), Patent Office Rules of Practice in Patent Cases (1968)
Perhaps in rejecting the claims "as having no novelty other than mental steps," the examiner meant to also include either or both 35 U.S.C. § 102 and 35 U.S.C. § 112 in addition to 35 U.S.C. § 101. See In re Prater, 56 CCPA ___, 415 F.2d 1393 (1969), in which the same primary examiner relied upon all three sections in a rejection similarly worded. Nevertheless, since 35 U.S.C. § 101 is the only section mentioned by the examiner, sections 102 and 112 are apparently not before us.
See n. 8, supra