Caterpillar Inc v Kozo Miyake

[2000] APO 3

OFFICIAL NOTICE

DECISION OF A DELEGATE OF THE COMMISSIONER OF PATENTS

Application  :          Nos. 651768 in the name of CATERPILLAR INC

Title:          Integrated Vehicle Positioning System, Apparatus and Method

Action:          Opposition under section 59 (Patents Act 1990) by KOZO MIYAKE

Decision:          Issued            .

Abstract

Extensive amendment of the claims after service of evidence in support. Opposition not withdrawn, but opponent not represented at the hearing, and no written submissions filed either. A primary document relied on was published before a 'learned society' and is protected by section 24(1)(a) and regulations 2.2(c) and 2.3(1)(b)(ii) of the Act. The evidence does not establish the common general knowledge in the relevant field, which is the control of autonomous vehicles. Section 40 particulars not followed by evidence or submissions, but not considered to be now applicable. Opposition wholly unsuccessful. Award of costs considered. Application to proceed to sealing.

PATENTS ACT 1990

DECISION OF A DELEGATE OF THE COMMISSIONER OF PATENTS

Re:Patent Application No. 651768 by CATERPILLAR INC and opposition thereto by KOZO MIYAKE

BACKGROUND

Patent application 651768 (the application) was filed as 46042/93 on 2 September 1993 in the name of CATERPILLAR INC (Caterpillar). The application is a divisional of patent 642638, from which it claims an earliest priority date of 11 December 1989, and it was advertised accepted on 28 July 1994. As the application was filed after the commencement date of the Patents Act 1990 the transitional provisions of that Act do not apply.

On 27 October 1994 a notice of opposition was lodged by KOZO MIYAKE (Miyake), followed by a statement of grounds and particulars on 27 January 1995.  The grounds relied on are novelty, obviousness, manner of manufacture, and clarity.  The evidence in support was served on 27 April 1995.  Requests to amend the statement of grounds and particulars were lodged on 19th and 27th April 1995.  These were opposed, and in a decision dated 14 December 1995 only the former was allowed.  Caterpillar's evidence in answer was served on 14 March 1996.

Another preliminary issue which went to hearing was Caterpillar's opposition to Miyake's request for an extension of time in which to serve a copy of a notice of intention to serve evidence in reply.  In a decision dated 16 October 1996 the extension was refused, but a short time was allowed for evidence at hand to be served.  This was done on 21 October 1996.

Section 104 proposals to amend the specification were filed on 15 April 1996 and 8 May 1997 and allowance was advertised on 15 January 1998 and 28 May 1998 respectively.  The amendments substantially narrowed the scope of the claims.

A hearing was held in Canberra on 11 November 1999.  Caterpillar was represented by Mr Douglas C Carter, patent attorney of Carter Smith & Beadle, Melbourne.  Miyake did not appear at all and filed no written submissions.

THE SPECIFICATION

The invention relates to the navigational control of an autonomous land-based vehicle.  In the specification 'autonomous' means that the vehicle is operated with no or substantially no human involvement.  A particular application mentioned is the operation of a mining vehicle, which can then be operated round the clock, in polluted or otherwise hazardous environments, or in hostile weather conditions, without risk to human life.

The system uses a quintic polynomial method to replan a simple continuous path from a present path back to a desired path in some look-ahead distance, and computes a steering angle corresponding to the part of the replanned path to be followed for the next time interval.  The 'look-ahead distance' is the automated equivalent of the distance in front of a manually driven vehicle which the driver looks.  A 'quintic polynomial' is a polynomial of the fifth degree, ie a polynomial in which the highest power term, or sum of the powers of a single term, is 5.

Another feature of the invention is that delays by system components in responding to commands is anticipated by setting the commands in advance.

However the specification is long, containing 58 pages of description and 85 sheets of drawings, and is mathematically complex.  Reflecting the existence of a considerable number of related applications and patents, much of it is not particularly relevant to the invention presently claimed, and I do not think it is useful for the present purpose to go further into it.

The specification concludes with 14 claims.  Claim 1 is the broadest claim and reads as follows; there is also a corresponding method claim.

"1. A system for navigational control of an autonomous, surface based vehicle of the type having an instantaneous center to which all motions can be referenced, and which uses wheels or tracks or a combination thereof and which exhibits either oversteer or understeer characteristics, comprising:

(1) first means to calculate a steering command to control steering of the vehicle, said  first means compensating the steering control for dynamic effects of the vehicle using feedback compensation using a quintic polynomial for replanning a continuous path from its actual position, heading and steering angle to rejoin a desired path; and

(2) second means using feedforward compensation to compensate for latency of the steering command, slow system dynamic response and vehicle dynamic characteristics."

EVIDENCE

From Miyake

· Declaration by Phillip McKerrow with exhibits PJM1 to PJM5.  Dr McKerrow is an electrical engineer, and at the time of his declaration he was employed as Senior Lecturer in the Computer Science Department of Wollongong University.  His background shows extensive experience in the fields of control systems, robotics, and autonomous vehicles.  The exhibits are extracts from texts and conference papers, principally concerning robotics.

· Declaration by Alex Zelinsky with exhibits AZ1 and AZ2.  Dr Zelinsky is employed in the Department of Computer Science at the University of Wollongong.  The exhibits are both conference papers relating to autonomous robots.

· Declaration by Michael A E Martin with exhibits MAEM1 to MAEM9.  Mr Martin is Miyake's patent attorney in this matter.  The exhibits are mainly some conference papers and JP patent specifications with translations.
· Declaration by Tony Cargnelutti, Librarian at the UNSW, going to the publication dates of various documents.

· Second declaration by Dr McKerrow, replying to Caterpillar's evidence in answer.

I note that exhibits PJM4, PJM5, and AZ2 cannot be considered as a result of the decision concerning amendment of the statement of grounds and particulars, and also that PJM1-3 and MAEM4 are not listed as particulars in that statement.

From Caterpillar

· Declaration by James  Philip Trevelyan, Associate Professor with the Department of Mechanical Engineering at the University of Western Australia.  Professor Trevelyan has also had extensive experience with robotics.

· Declaration by Tracey Joan Hendy with exhibit TJH1.  Ms Hendy is one of Caterpillar's patent attorneys in this opposition.  The exhibit relates to the various activities and other matters of the Institute of Electrical and Electronics Engineers (IEEE).

· Affidavit by In Soo Ahn, also concerning the IEEE.

SUBMISSIONS

From Miyake

As I have said, Miyake was neither represented at the hearing nor made any written submissions.  In addition, the claims were substantially amended after the evidence in support was lodged.  Present claim 1 is closely equivalent to original claim 5, appended to original claim 4, when it was appended to original claim1.  The evidence in support also includes reference to some documents which were disallowed in an earlier decision, and documents not particularised.  Consequently much of the evidence in support, both in declarations and the exhibits, is no longer very relevant.  Dr McKerrow is the only one of Miyake's declarants to address the issues.  I will pick up what he says at paragraph 100 of his first declaration, where he is referring to original claim 4.

· In a system such as that claimed, the number of constraints in the system determines the power of the polynomial used.  It is not a matter of choice.

· Specifically, to define a trajectory for a given set of start and finish positions, start and finish velocities, and start and finish accelerations, requires a quintic polynomial.  This is common general knowledge (cgk), and is found in robotics textbooks.  See the textbook extracts supplied as PJM2 and PJM3.

· The problem of tracking the path of a robot arm is no different from that of tracking the path of an autonomous vehicle.  It is obvious to apply this solution to the path tracking of an autonomous vehicle.

· Regarding claim 5 (present claim 1), the paper by Singh and Shin (AZ1) discloses using a quintic polynomial to describe the path between the current path and the desired path.  The point at which the replanned path converges with the desired path is the look ahead distance.  The method described in the description of the present application is exactly the same as disclosed by Singh and Shin.

The following is from McKerrow's second declaration, which is evidence in reply, and the claims are as amended.

· Professor Trevelyan (Caterpillar's principal declarant) does not actually say that he considers the claimed invention novel and inventive over all the prior art, including AZ1.

· In considering this paper Trevelyan appears to discount it because it specifically states that dynamic effects are not considered.  However this indicates that they were aware of them, as is also shown by stating that in the absence of taking them into consideration "stability and convergence cannot be guaranteed".  In effect Singh & Shin have kept their initial work simple to demonstrate that it works, and as indicated at the last sentence "future work will attempt to incorporate vehicle physics into the control scheme", this an obvious thing to do.

· Trevelyan does not seriously dispute that the use of feedforward compensation for latency effects and dynamic response errors, the use of feedback compensation, and the use of quintic polynomials, are all cgk at the priority date.

· It is inherent that the steering of an autonomous vehicle must be controlled.  The fact that the claims specifically mention steering does not add inventiveness.

· The paper by Lee (exhibit MAEM1) discloses all the features of the claims, including calculating the steering commands in advance (claims 3 and 9), except the quintic polynomial.

· The paper by Cheok et al (MAEM2) also discloses all the features claimed, except for the quintic polynomial.

· But the use of quintic polynomials to specify trajectories in terms of the desired position, velocity and acceleration of both the beginning and end of path segments is cgk.  Therefore at least claims 1 and 7 are obvious in view of cgk alone, cgk together with AZ1, or cgk together with MAEM1.

From Caterpillar

In the following I summarise what seem to me to be the most important of Caterpillar's submissions.

· The exhibit AZ1 is a paper given by two of the inventors (Singh &Shin) of the present invention at the IEEE/RSJ International Workshop on Intelligent Robots and Systems on 4-6 September 1989 in Japan. The presentation of this paper is protected by the provisions of section 24(1)(a) of the Patents Act 1990, and regulation 2.2(2)(c). Section 24(1)(a) provides that, for the purpose of deciding whether an invention is novel or involves an inventive step, any information made publicly available in the prescribed circumstances set out in regulation 2.2, by or with the consent of the nominated person or predecessor in title, must be disregarded if a patent application is made within the prescribed time set out in regulation 2.3.

In the present case, the patent application from which the subject application claims priority was filed in the US on 11 December 1989, which is well within the six months allowed by regulation 2.3(1)(b).

In his declaration Dr Zelinsky states that he attended the above-mentioned "scientific conference held in Tsubaka, Japan on 4-6 September 1989", where he received a copy of the proceedings of the conference, including a copy of the Singh & Shin paper.

The declarations by Hendy and Soo Ahn, which set out the structure and activities of the IEEE, show that it is a 'learned society' within the meaning of the Act, and the disclosure must be disregarded in this proceeding.  It is also noted that the evidence in reply does not address this issue.

· In any case, AZ1 doe not disclose the full combination of features claimed.  This paper omits any consideration of the dynamic effects of the vehicle, as is conceded in the evidence in reply.  It is suggested there that because Singh & Shin twice comment that dynamic effects are not considered, that this indicates that they need to be considered, and considered in the way which is now claimed.  An alternate explanation may well be that while Singh & Shin were aware that dynamic effects were not considered, it was other of the 23 in all listed inventors in the application who, in fact, contributed the ideas associated with the use of feed forward compensation for the dynamic effects of the vehicle.

· In view of the amended claims, Miyake now evidently relies only on the ground of obviousness.  However there is a complete failure to show what was the cgk in Australia at the priority date, or that any of the items in evidence were cgk.  As obviousness must be assessed against the background of the cgk, this ground must fail.

· As regards the other grounds, there is no evidence in support nor substantive submission in respect of them, and it is put that they are without merit.

· Full costs are claimed for this opposition, notwithstanding the fact that the claims were amended as a result.  Miyake has, by reason of not appearing at the hearing and not making any submissions, tacitly agreed that the amended claims avoid the grounds for the opposition.

DECISION

Novelty

It is settled that the test for novelty is the reverse infringement test, in which, to establish want of novelty in a claim, each and every one of the essential features of the claim must be found in the allegedly infringing article.  See Meyers Taylor Pty Ltd v Vicarr Industries Ltd (1977) CLR 228 and Rodi & Weinenberger AG Henry Showell Ltd (1969) RPC 367.

I can see no reason to regard any of the features of the broadest claim as inessential, and there has been no submission otherwise.

The Shingh & Shin paper (AZ1)

Regarding the question of the publication of AZ1, the parts of the Act and Regulations which are relevant to the stated facts of this case are as follows.

"24 Validity not affected by certain publication or use

(1) For the purpose of deciding whether an invention is novel or involves an

inventive step, the person making the decision must disregard:

(a) any information made publicly available, through any publication
or use of the invention in the prescribed circumstances, by or with
the consent of the nominated person or patentee……
but only if a patent application for the invention is made within the
                prescribed period."

"Publication or use: prescribed circumstances and periods

2.2 (1) In this regulation:……
           (2) For the purposes of paragraph 24 (1) (a) of the Act……the following
                circumstances are prescribed:
           ………………
           (c) the publication of the invention in a paper written by the inventor and:
  (i)  read before a learned society; or
  (ii) published with the inventor's consent by or on behalf of a learned society;"

"Prescribed period: publication or use affecting validity

2.3 (1) For the purposes of subsection 24 (1) of the Act……in the case of
  information of the kind referred to in paragraph 24 (1)(a) of the Act; the
  prescribed period is:
           ………………
           (b) in the case of the circumstances mentioned in paragraph 2.2 (2)(c):
           ………………
  (ii) in any other case - 6 months after the first reading or publication;……"

On the face of it AZ1 appears to be protected by these provisions.  There is no submission from Miyake on this issue, and in fact it was only expressly raised at the hearing, though the evidence in answer does include material concerning the nature of the IEEE organisation.  The only question appears to be whether the IEEE qualifies as a 'learned society'.  The following is paragraph 3.4.6.2.1 of the Australian Patent Office Manual of Practice and Procedure Volume 2, citing a decision of the UK Patent Office in similar circumstances.

"There is no complete definition of what constitutes a 'learned society'.  However, the judgement in Ralph M Parsons Co (Beavon's) Application (1978) FSR 226, gives some guidance. This judgement suggests that 'learned society' is apt to describe any non-commercial body of persons seeking to promote and organise the development of specific subjects by the provision of a forum for the exchange and discussion of ideas and the dissemination of information, usually through the publication of its proceedings."

Guided by this and the evidence filed in answer, I consider that the IEEE is a learned society within the meaning of the Act, and the publication of AZ1 is therefore protected in respect of both novelty and inventive step.  Also, I agree with Caterpillar's submission in respect of novelty that in any case, in not dealing with vehicle dynamics, AZ1 does not disclose all of the features claimed.  Consequently I will pay no further attention to this document.

Other documents cited

It seems clear to me that none of the other documents disclose every one of the features claimed.  The closest are probably the papers by Lee (MAEM1) and Cheok et al (MAEM2), and neither of them disclose the use of a quintic polynomial.  Consequently I find that the ground of lack of novelty has not been made out.  And indeed, from the evidence in reply (though this may have been cut short), it appears that this ground has been abandoned.

Inventive step

The question of inventive step or obviousness has to be decided against the background of the common general knowledge (cgk) in the art, in the patent area, at the priority date. Also, cgk can be taken with a prior disclosure, so long as a person skilled in the art could be reasonably expected to have ascertained, understood, and regarded it as relevant (section 7(2) and (3), Patents Act 1990). I consider that the relevant field in this case relates to the control of autonomous vehicles, not robot machinery (perhaps unless mobile), and not process control.

However, although the parties' principal declarants are clearly experts in this field, and are independent, there is little if any agreement as to what was the cgk, specifically in the matter of the control of autonomous vehicles, at the priority date.

The onus is on the opponent to make its case.  In circumstances such as this, where the experts disagree, I consider that more is required than unsupported statements by Dr McKerrow that something is cgk, before I can accept it.  In discussing cgk, McKerrow also refers (in the evidence in reply) to documents on file but which have been excluded or are not particularised.  It was submitted that these documents can nevertheless be used as evidence of techniques which are cgk.

But even if this is so I do not think it is helpful to the opposition case.  Some of the documents are not concerned with the control of autonomous vehicles, some are research papers, published within a year or two of the priority date, and most are published overseas.  In my opinion they do not provide a convincing basis for concluding what the cgk is, in the field of autonomous vehicle control, at the priority date, in Australia.

There is also no evidence, in my opinion, that any of the particularised documents were cgk, and it seems to me perhaps unlikely that they would have been, since they are all either papers presented at some conference, a published collection of papers, or JP patent specifications.

In my opinion the evidence does not establish what was the cgk in the relevant art at the priority date in Australia, and it is also appropriate, I believe, to give the applicant the benefit of any doubt in the matter.  I conclude that the ground of obviousness fails.

Section 40

There are a large number of particulars under this ground, mainly clarity and fair basis of the claims, and they have not been followed up by any evidence or submissions.  Also, they are directed to the specification as accepted.  However I have considered them briefly.

In my opinion there are no important section 40 defects in the claims, and while the description rather untidily reflects the large number of related, divisional, and divisional of divisional applications, I do not consider that it fails to satisfy the requirements of section 40.

CONCLUSION

I have decided that none of the grounds of opposition have been made out.  Consequently I direct that the application proceed to sealing, subject to there being no appeal.

COSTS

In proceedings before the Commissioner costs normally follow the event, but all relevant factors should be taken into account in making the award.  In this case Caterpillar's own declarant expresses the view that the accepted claims were too broad in view of the prior art contained in the evidence in support.  The claims were subsequently substantially amended, and I have found that they are now in order.  In these circumstances it would be common practice to award costs against the applicant up until the time the amendments were allowed.

However in this case Caterpillar has claimed full costs in relation to the opposition, saying that Miyake's non-appearance at the hearing and failure to file any submissions in support of the opposition is a tacit recognition that the amended claims avoid the grounds of opposition.

I do consider that Miyake's failure to either appear or make written submissions has not been helpful in the opposition process; another option would have been to withdraw the opposition.  Nonetheless the opponent was entitled to proceed in the way chosen, but given the outcome, must bear costs arising after allowance of the amendments.  However I do not think it appropriate that Miyake should bear the other costs as well.  This was a serious and effective opposition.  Had the opposition proceeded without the claims being amended the result would likely have been different.

I consider that Caterpillar's decision to appear at a hearing, and to be represented at it by a single patent attorney, is a reasonable one.  Consequently I award costs up until the time of allowance of the second set of amendments against the applicant Caterpillar Inc, and thereafter against the opponent Kozo Miyake.

RG Tolhurst
Delegate of the Commissioner of Patents

Patent attorneys for the applicant  :  Carter Smith & Beadle, Melbourne

Patent attorneys for the opponent   :  Chrysiliou Moore Martin