Mr Greg Fendis, CEO of the three propriety limited companies: Austcom Pty Ltd, Biofilter Pty Ltd (Biofilter) and GOGlobal Pty Ltd;

Mr Sachlan Fraval, a chartered accountant and director of the applicant and the father of Mr Sofiaan Fraval; and

Dr Hadrian Fraval, the Managing Director of Rofin Australia Pty Ltd (Rofin), the brother of Mr Sachlan Fraval and the uncle of Mr Sofiaan Fraval.;

1. On behalf of the respondent, the Tribunal heard from Professor PMM Wagner, Professor of Computing, and the Director of the National Centre for Biometric Studies established in 2005 at Canberra University.
2. Mr Sachlan Fraval appeared for the applicant, assisted on occasions by Dr Hadrian Fraval and Mr Fendis (the latter cross-examining Professor Wagner). Mr Rebikoff, of counsel, represented the respondent. At the conclusion of the oral evidence, the applicant was given the opportunity to submit further questions in writing to Professor Wagner to answer. Difficulties ensued with respect to the relevance of the questions, which the applicant reformulated into seven questions and resubmitted with the copy of a patent application. Professor Wagner provided a further report in answer to the questions on the 16 July 2010. Subsequently, the parties provided, and exchanged, written submissions, and the respondent also provided the "T Documents" pursuant to s 37 of the Administrative Appeals Tribunal Act 1975.

1. The ITA Act provides tax deductions for eligible companies which undertake research and development activities. Section 73B (1) of the ITA Act defines "research and development activities" in two interrelated parts. If a project is found to involve systematic, investigative and experimental activities, then consideration is to be given to other activities carried on for a purpose directly related to them. Section 73B(1) of the ITA Act provides,

(a) systematic, investigative and experimental activities that involve innovation or high levels of technical risk and are carried on for the purpose of:

(i) acquiring new knowledge (whether or not that knowledge will have a specific practical application); or

(ii) creating new or improved materials, products, devices, processes or services; or

(b) other activities that are carried on for a purpose directly related to the carrying on of activities of the kind referred to in paragraph (a).

1. Section 73B (2B) of ITA Act excludes certain activities from constituting "research and development activities". It provides:

[f]or the purposes of the definition of research and development activities in subsection (1):

(a) activities are not taken to involve innovation unless they involve an appreciable element of novelty; and

(b) activities are not taken to involve high levels of technical risk unless:

(i) the probability of obtaining the technical or scientific outcome of the activities cannot be known or determined in advance on the basis of current knowledge or experience; and

(ii) the uncertainty of obtaining the outcome can be removed only through a program of systematic, investigative and experimental activities in which scientific method has been applied, in a systematic progression of work (based on principles of physical, biological, chemical, medical, engineering or computer sciences) from hypothesis to experiment, observation and evaluation, followed by logical conclusions.

1. Since this review involves the development of computer software, s 73B (2A) of the ITA Act is also relevant. It provides that:

activities carried on by or on behalf of an eligible company by way of the development of computer software shall not be taken to be systematic, investigative and experimental activities unless the computer software is developed for the purpose, or for purposes that include the purpose, of sale, rent, licence, hire or lease to 2 or more non-associates of the company (counting a non-associate of the company and the associates of such a non-associate together as one person). ["Multiple Sale Criterion"].

1. The aim of the applicant's R and D plan, dated 1 July 2007, was the development of a replacement for the current carparking methodology for identifying customers wishing to access carparks. A biometric system was envisaged, which utilised either a camera or a fingerprint scanner to recognise the customer upon ingress and also to calculate the amount owing on egress. This would eliminate the need for a carparking customer to retain and use a card or token. It was also anticipated that a peripheral device, which integrated the proposed identity recognition with optical character recognition of registration plate letters and figures, could provide additional security[2].
2. On behalf of the applicant, there have been at least three plans prepared, which are said to constitute the core development activity[3] It is that one set out in the witness statement of Mr Sachlan Fraval and prepared for this hearing, which is the most relevant. It reads as follows:

Segment A 1. CAR PARKING APPLICATION

2. CAR PARKING EQUIPMENT

2.1 CDS

2.2 Other

3. CUSTOMER DATA

3.1 Registration

3.2 Authentication

4. PHYSICAL PARAMETERS

4.1 Physical Devices

4.2 Environment

4.3 Other

5. DATA

5.1 Collection

5.2 Storage

5.3 Usage

6. BIOMETRICS

6.1 Algorithm

6.1.1 Fingerprint

6.1.1.1 Rofin

6.1.2 Iris

6.1.3 FACE

6.1.3.1 Biofilter

6.2 Application Integration

Segment B 7. NON-ATTENDANCE / REMOTE DECISION MAKING

8. ARTIFICAL INTELLIGENCE

Segment C 9. BUSINESS RULES

10. ALGORITHM

1. It is claimed that activities 7-10 are the core activities claimed for the purposes of s 73B(1)(a) of ITA Act, and that numbers 1-5 are the directly related activities. Mr Sachlan Fraval told the Tribunal that no claim is made for activity 6. This was emphasised in a subsequent submission under the heading, "Explanation of Algorithm Development", conveyed from the applicant to the respondent under cover of a letter dated 19 August 2010. It was claimed that the applicant had entered into a joint agreement with a company called Biofilter to apply facial recognition technology to carparking applications[4]. Biofilter was a company established in approximately 2000 by Rofin[5].
2. The first thing to determine is the activities in which the applicant was engaged. The usual way in which activities are determined is by reference to records. Usually, records will be in a documentary, or computer generated time sequenced, form and consist of R and D plans, outcomes of test results, revision control records, minutes of meetings at which development of the project was discussed, invoices for costs incurred and the like. A starting point is to examine the application forms for registration of R and D activities. The first claim form, dated 20 September 2004, nominated the title for the project as, "Facial Recognition ASP System for Car Parks", and it lists the following as the activities undertaken in the 2003-2004 year:

[a]cquired licence to use facial recognition and infrastructure technology to integrate facial recognition into an ASP system to control at an account level or at a cash or credit card securely the customers entering a pay car park at the boom gates[6].

1. The 2003-2004 plan nominated the development of a biometric system using a development toolkit provided by Biometriq_[7]. The plan described Biometriq as a specialist biometric company[8], and the evidence established it as being located in Hong Kong[9]. It was proposed to replace the then current usage of carpark customer recognition utilising a card or token by a system using photographic or fingerprint scanner recognition[10]. It appears from the oral evidence that voice recognition was also considered, but it was rejected early in the process. Utilisation of a biometric methodology (or a combination of biometric methodologies) was aimed at establishing the personal identification of the user. The plan also stated that Rofin was to provide "biometric expertise"[11]_.
2. In a letter dated 7 March 2004, the respondent posed a number of questions to the applicant.[12] Answers were sent in a response, dated 19 July 2005_[13]_. While it is evident that no final decision had been made about the form of the biometric application to be used, iris and fingerprint technology had been rejected. In respect of "innovation or high levels of technical risk", appears the comment:

[v]ery high level of technical risk was incurred in the project, evidence of which was inability to integrate successfully the iris and fingerprint technologies which cost a significant amount in the development process to date, and will not be used in the final product[14].

1. There were two facets of the work to be undertaken. One involved the identification of the most appropriate biometric medium to be used in identity recognition, whether face, voice, iris or fingerprint. The other facet was the hooking (or connection) of the information to the software used in a carpark environment, for example, to operate the boom gates and to calculate of the time spent and/or money to be paid on exit.
2. The R and D was claimed to be carried out by Mr Sofiaan Fraval, a director of the applicant. He holds a Bachelor of Digital Systems from Monash University, and his occupation is an embedded systems engineer. In his evidence to the Tribunal, Mr Sofiaan Fraval said that he was the main R and D technologist for the project[15]. He said that the time he spent working on the project, the subject of this application, was about 40 hours a week. Mr Sofiaan Fraval agreed with Mr Rebikoff that, in a simplified form, once the biometric medium had been selected, there were three basic elements to the project. They are as follows: (a) identifying existing biometric facial recognition software; (b) car parking system purchased being hardware (boomgates) and some control software from CDS Worldwide Pty Ltd (CDS), which Mr Sofiaan Fraval described as being widely used in carparks;[16] and (c) online billing and payment system[17]. The aim of the project was to develop software which would link these three elements.
3. After considering the fingerprint, voice, iris and facial recognition options, the last one was chosen as the best method of authenticating carpark users. Mr Fendis explained that the utilisation of fingerprint evidence was rejected, because, at least in the mind of the public, it was associated with criminal procedures. He also said that in any event, a fingerprint provided only 17 points of identification, whereas facial imaging returned 4000 identification points.[18]
4. Mr Sofiaan Fraval told the Tribunal that there was no record available of the work carried out. Sometime in early 2006, the hard drive on his computer, on which the records were retained, had been damaged, and he had been unable to afford to retrieve the information. No back up of the computerised information was in place,[19] including the revision control system which identified changes made to systems after conducting trial and error tests[20].
5. Mr Sofiann Fraval agreed that it was a negligent omission on his part not to have backup procedures[21]. There were no time sheets maintained[22]. Note pad entries were unable to be produced as he had either lost or destroyed them[23]. It is not, apparently, the case that the information contained on the hard drive has been irretrievably lost. Mr Sofiaan Fraval said that without the guarantee of succeeding in the R and D claim made for $1.36 million in this review, it would not be worth expending up to $5,000 to enable the information to be retrieved[24]. The absence of any records presents difficulties for the determination of this application, including as to when the project commenced - Mr Sofiaan Fraval found difficulty in recalling when the project proceeded from mere discussion to a commencement phase[25].
6. In his first report, Professor Wagner commented that "[t]o lose one's life's work, as this appears to have been, without any backup would be inexcusable for an ordinary person - for a qualified computer engineer such gross negligence is all but incomprehensible"[26]. The Professor told the Tribunal that it could be expected, when conducting work of this kind, that there would be backup undertaken at least once a week[27]. That this was not done he found to be "astounding"[28].
7. Unfortunately, when cross-examined, Mr Sofiaan Fraval was unable to provide particularisation of any testing carried out as the project advanced due to the lack of contemporaneous records[29]. This presents a major problem as, according to Mr Fendis' evidence, the applicant's

software development was based on an iterative and outcome based process. We wanted a predictable schedule and outcome that could be used as a fundamental starting point for other software developers with lower skills[30].

1. The Tribunal finds it difficult to comprehend how that could be achieved without the development of conjunctional testing criteria. The evidence of Mr Sofiaan Fraval on this issue was vague to the point that he said he could not recall what was on the testing checklist. He stated that testing consisted of listing "the positives and negatives"[31]. At another point, he told the Tribunal that the development of a biometric algorithm designed to permit identification of a person's image was investigated, but it proved to be unsuccessful[32]. To the extent that his work on the project involved this aspect, the Tribunal has discounted it as Mr Sachlan Fraval opened by saying that it did not form part of the claim in this application. The problem is that the Tribunal is unable to assess what time Mr Sofiaan Fraval worked on this aspect and, if so, to what extent, if any, it formed part of the claim.
2. It was Mr Sofiaan Fraval's evidence, in his statement, that he worked fulltime on the applicant's project[33]. However, to the Tribunal, he said that he was engaged by CDS and worked on the applicant's project in his spare time and on weekends, estimating that he spent 40 hours a week on the project[34]. Mr Sofiaan Fravel told the Tribunal that he was engaged as a contractor to CDS, where he worked developing car locking and security patrolling devices. The implication conveyed by his evidence is that, since the respondent had not funded the R and D and, consequentially, he had been paid nothing for his work in the project, he was obliged to leave Australia to advance his career[35]. In light of his oral evidence, that he was a working as a paid contractor to CDS and that he was not engaged fulltime on the applicant's project, the Tribunal is unable to accept the implication conveyed and is satisfied that, contrary to what he said in his statement, he did not work on the project fulltime.
3. However, the Tribunal accepts Mr Sofiaan Fraval's evidence that he spent considerable time on the HZXD project. He stated that:

[t]he work was undertaken to prepare a proprietary system for the facial recognition interface after systematically investigating other technologies that provided biometric security enhancements over the traditional pay and park systems.[36]

1. Mr Sofiaan Fraval told the Tribunal that he engaged in modifying the CDS developed iPark software utilising a prepared application programming interface system (API) obtained from Biometriq. The T Documents contain the modified version of the 400 page document entitled, HZXD "...Biometric Car park application: Software Listing"[37]. Some, at least, of the modification seems to have involved substituting HZXD's name for that of CDS, which he described as "it was a big deal to do that"[38]; some functions involved rewriting information contained in a different computer language[39]; some developmental work to ensure implementation of the existing codes[40]; but overall to achieve what he accepted as being "...hooks to enable connectivity with different technologies and algorithms"[41].
2. It was also Mr Sofiaan Fraval's evidence that he alone was responsible for driving the project[42]. However, in his evidence Dr Hadrian Fraval maintained that Mr Sofiaan Fraval was in reality "...one of a team of people"[43]. At the conclusion of his oral evidence, Mr Sofiaan Fraval said that the work on items 8, 9 and 10 of the plan (referred to in paragraph 10 above) were undertaken by Mr Fendis, but he was unaware of which capacity Mr Fendis was acting[44].
3. Mr Fendis provided a statement, dated 12 February 2009[45]. It consisted of selected word for word extracts from a previous letter, dated 28 September 2007, sent by the applicant to the respondent[46]. Mr Fendis claimed that it may have been the situation that his words in an earlier unidentified document were used by someone else in his/her document[47]. Mr Fendis denied that he had adopted the content of his statement without considering whether that content reflected the activities undertaken in the claimed R and D carried out by the applicant[48]. Given the date of the letter and the date Mr Fendis maintained he had prepared his statement, the Tribunal is satisfied that Mr Fendis' statement is constituted by extracts copied from the prior letter. However, the Tribunal recognises that it does not follow that, because this is so, the contents of the statement do not represent the views held by him.
4. It was Mr Fendis' evidence that while he agreed GOGlobal had developed a generic host application program, which permitted other software applications to link to its online billing and credit card authorisation program, he was unable to say if HZXD had accessed the program[49]. However, Mr Fendis confirmed that Biofilter, pursuant to a joint venture arrangement with the applicant, supplied the applicant with facial recognition software[50]. It was Mr Fendis' evidence that Biofilter "generally" developed the biometric applications as part of the joint venture partnerships in which it engaged[51]. In respect of the facial software development, other than the joint venture, Mr Fendis said that Biofilter operated independently from the applicant and had no involvement in the day to day R and D operations undertaken by the applicant[52].
5. Dr Hadrian Fraval told the Tribunal that Mr Sofiaan Fraval had undertaken work for the project both at CDS and Biofilter[53]. It was his evidence that CDS, Biofilter and Rofin were all part of the project, and that he was involved from the outset. Dr Fraval said that two of Rofin's personnel had been involved part-time on the project[54], as well as other contractors[55], and that Rofin billed the applicant for approximately $75,000 for work carried out[56]. No request had been made by the applicant to Rofin to produce any records it may have on the project[57]. Dr Fraval said that his input was to advise the applicant of the pitfalls of using the then available biometric technology[58], and that the Rofin staff were involved at the stage when fingerprint biometrics was being considered[59]. While he could recall attending meetings connected with the project, Dr Fraval was unable to recall if minutes were kept before confirming communication would mainly have been by way of discussion[60].
6. Dr Fraval told the Tribunal that, "...the absolute the [sic] nub of this whole project is artificial intelligence"[61]. He described "artificial intelligence" as consideration of the data from an algorithm to determine ways it could be changed in order to get closer to 100% accuracy in biometric recognition[62]. Dr Fraval said that Rofin had now taken over the project.
7. It was Professor Wagner's evidence that, while he acknowledged there was plenty of work to engage a programmer in writing the programs necessary to bring the project to fruition, the work did not involve innovation or technical risk[63]. He said that standard transmission and internet protocols existed at that time, which could be utilised to prepare the software connections[64]. The Professor's evidence coincided with that of the applicant that using facial recognition would never be 100% secure. Programming required the determination of a threshold of risk in the use of the biometric settings. However this, in the Professor's view, while it may involve experimentation, did not involve innovation[65].

1. The evidence establishes to the Tribunal's satisfaction that Mr Sofiaan Fraval was the person responsible for undertaking the R and D in respect of the claim, and that his father, Mr Sachlan Fraval, was responsible for the business organisation of the applicant. However, in view of Dr Hadrian Fraval's evidence, the Tribunal cannot be satisfied that Mr Sofiaan Fraval was the only person working on the project, nor that HZXD was the only company involved in carrying out the R and D connected with the program.
2. The Tribunal is satisfied that the initial work was carried out in the 2003-2004 year to identify the form of biometric technology which would be suitable for adaptation in the project. This was carried out by examining various applications including fingerprint, iris, voice and facial recognition. While ultimately facial recognition was chosen as the preferred biometric medium, it is unknown what experiments were carried out in order to reach that decision. In the absence of that pertinent information, the Tribunal is unable to conclude that whatever work was done was either innovative or involved high levels of technical risk.
3. The 400 page document entitled HZXD "...Biometric Car park application: Software Listing" was a software program originally issued by CDS. While the Tribunal is satisfied that Mr Sofiaan Fraval modified the program for at least part of the work relevant to the application, it is unable to determine the content or extent of that modification. The lack of supporting evidence leaves the Tribunal unable to calculate what portion, if any, of the application should be accepted as involving innovation or a high degree of technical risk. The Tribunal is satisfied that the applicant had ample opportunity, over a lengthy period of time, to locate and to provide more detailed evidence, but none was forthcoming.
4. The same is true of other software programs obtained by the applicant and modified by Mr Sofiaan Fraval. For example, the kit obtained from GOGlobal which was to be modified for utilisation in the payment process. The Tribunal is satisfied, as Dr Hadrian Fraval confirmed[66], that no software codes for application in the project were developed.
5. Since there were no contemporaneous records of what work was undertaken, and when it was conducted including the results of any claimed trial and error processes, and since the claimed R and D has not resulted in any identifiable product, the Tribunal is satisfied that it is impossible to assess what was done constituted systematic, innovative or experimental work or involved technical risk.
6. In any event, the evidence of Professor Wagner is to the effect that, while programming work was carried out, it did not even at the time of the claim commencing in 2002 involve "innovation" or "high technical risk". The Tribunal accepts that there is a distinction between conducting trial and error experiments, as was claimed by the applicant to be undertaken during the course of the project, and innovation. The evidence points to the former being undertaken to determine whether, in a system which can never be risk free, an acceptable level of risk can be established. The experiments determined whether one, two or more photographs of the individual's face needed to be taken to ensure as accurate as possible recognition or, if not, whether greater security could be ensured if facial recognition was combined with other modes of recognition (for example, car registration plates). This goes to the efficiency and practicality of the chosen method, but it does not evidence 'innovation'. That is, in the terms of s 73B (2B)(a) of ITA Act, the Tribunal cannot be satisfied that there is "an appreciable element of novelty". Nor does such testing involve a high level of technical risk as defined in s 73B (2B)(b)(ii) of ITA Act, as there is no evidence that any "scientific method has been applied, in a systematic progression of work (based on principles of physical, biological, chemical, medical, engineering or computer sciences)".
7. Mr Fendis' evidence that the carpark project was solely worked on by the applicant is inconsistent with Mr Sofiaan Fraval's evidence that it was Mr Fendis, who was responsible for the development of items 8, 9 and 10 of the plan referred to in paragraph 10 above. The only evidence of Mr Fendis being involved on the project was a suggestion by Dr Hadrian Fraval that Biofilter the company of which Mr Fendis is the CEO may have been a contractor[67]. If Mr Fendis, or Biofilter, carried out the work nominated by Mr Sofiaan Fraval, then the Tribunal is unable to conclude that it was carried out by the applicant. Even accepting Mr Sofiaan Fraval's evidence that Mr Fendis, whether in his capacity as the CEO of Biofilter or in a personal or some other capacity, did some work on the project, the Tribunal is unable to determine what that work involved. In as far as this application is concerned, the Tribunal is left unable to be satisfied that any work undertaken for items 8, 9 and 10 of the plan can be described as "being systematic, investigative or experimental". Nor can the Tribunal be satisfied that any such work involved "innovation or high levels of technical risk".
8. The Tribunal was urged by Mr Sachlan Fraval to look at the project as a whole to determine whether it was innovative or involved a high level of technical risk. In looking at the project as a whole, the Tribunal understands the project to consist of the adaptation of interfaces to connect pre-existing identification systems to replace card or token systems used for carpark management. Without some more defined descriptions of the work actually carried out as R and D, in the Tribunal's view, the project as a whole does not demonstrate the characteristics of R and D as required by the ITA Act.
9. For the reasons given, the Tribunal affirms the decision under review.

I certify that the 40 preceding paragraphs are a true copy of the reasons for the decision herein of

For the Applicant Mr Sachlan Fraval, Dr H Fraval and

Solicitor for the Respondent Mallesons Stephen Jaques

[1] Income Tax Assessment Act 1936, ss 73B- 73Z and Industry Research and Development Act 1986, ss 39L and 39S (IR& D Act). The Australian Tax Office requests the respondent to issue a certificate pursuant to s 39L of the IR & D Act as to whether the activities in question constitute R and D for the purposes of s 73B of the ITA Act; s 39S of the IR & D Act provides for internal review of a decision as to whether activities constitute R & D.

[3] The first, dated 1 July 2003, T Documents, T 52; the second, dated 28 September 2007, T Documents, T 18; and the third is set out in the witness statement of Sachlan Fraval, dated 2 February 2010, Exhibit A1.

[18] Transcript, p 78 and T Documents, T 50 at p 303.

[23] Ibid, p 181 and Exhibit A4, p 1.

[37] T Documents, T 53.The full version was provided to the Tribunal pursuant to its Amended Direction, 28 July 2009.