REASONS FOR JUDGMENT
Jorré J.
Introduction
[1]
On January 27, 2011, the Minister of National Revenue
issued a reassessment for the 2009 taxation year. The Minister reduced the amount
of the scientific research and experimental development tax credit claimed by
the appellant for the following projects:
(a) 2007-01: Development of a new assembling process for motors
for dryers (this project started in 2007).
(b) 2007-02: Development of a new assembling process for heating
elements for dryers (this project started in 2007).
(c) 2009-01: Development of a new assembling process for control
panels using pull flow with electronic sequencing.
(d) 2009-02: High-speed synchronization for the application of
secondary components on a print component.
[2]
The vast majority of expenditures giving rise to
the credit claimed are salary expenditures incurred during testing on the
production line.
[3]
The Minister submits that the activities of the
appellant involved no scientific uncertainty, that they were not based on a systematic investigation or search that was carried out in a field
of science or technology and that they were merely routine activities
with no basic research, no applied research and no experimental development.
For these reasons, the Minister claims that he properly disallowed the
appellant’s claim.
[4]
There is no doubt that the purpose of the projects
in question was to increase production efficiency and, in one case, to ensure innovative production.
[5]
The appellant challenges the Minister’s decision
and submits that the projects were eligible for the credit claimed.
[6]
Section 248 of the Income Tax Act defines
“scientific research and experimental development” as follows:
“scientific research and experimental development” means systematic
investigation or search that is carried out in a field of science or technology
by means of experiment or analysis and that is
(a) basic
research, namely, work undertaken for the advancement of scientific knowledge
without a specific practical application in view,
(b) applied
research, namely, work undertaken for the advancement of scientific knowledge
with a specific practical application in view, or
(c) experimental
development, namely, work undertaken for the purpose of achieving technological
advancement for the purpose of creating new, or improving existing, materials,
devices, products or processes, including incremental improvements thereto,
and, in
applying this definition in respect of a taxpayer, includes
(d) work
undertaken by or on behalf of the taxpayer with respect to engineering, design,
operations research, mathematical analysis, computer programming, data
collection, testing or psychological research, where the work is commensurate
with the needs, and directly in support, of work described in paragraph (a), (b), or (c) that is undertaken in
Canada by or on behalf of the taxpayer,
but does
not include work with respect to
(e) market
research or sales promotion,
(f) quality
control or routine testing of materials, devices, products or processes,
(g) research
in the social sciences or the humanities,
(h) prospecting,
exploring or drilling for, or producing, minerals, petroleum or natural gas,
(i) the
commercial production of a new or improved material, device or product or the
commercial use of a new or improved process,
(j) style
changes, or
(k) routine data collection;
[7]
If we remove those parts which are not relevant
to this dispute, the definition reads as follows:
Systematic investigation or search that is
carried out in a field of science or technology by means of experiment or
analysis and that is
(a) . . .
(b) applied
research, namely, work undertaken for the advancement of scientific knowledge
with a specific practical application in view, or
(c) experimental
development, namely, work undertaken for the purpose of achieving technological
advancement for the purpose of creating new . . . devices . . . or processes,
including incremental improvements thereto,
and, in
applying this definition in respect of a taxpayer, includes
(d) work undertaken by or on behalf of the taxpayer with respect
to engineering, design, operations research, mathematical analysis, computer
programming, data collection, testing or psychological research, where the work
is commensurate with the needs, and directly in support, of work described in
paragraph (a), (b), or (c) that is undertaken in Canada by or on behalf of the taxpayer,
but does not include work with respect to
(e) . . .
(f) quality
control or routine testing of materials, devices, products or processes,
(g) . . .
(h) . . .
(i) the
commercial production of a new or improved material, device or product or the
commercial use of a new or improved process,
(j) . . .
(k) . . .
[8]
Ultimately, the central isssue is whether the
projects in question constitute:
(c) experimental
development, namely, work undertaken for the purpose of achieving technological
advancement for the purpose of creating new . . . devices . . . or processes,
including incremental improvements thereto,
[9]
At the heart of the controversy is whether it is
technological advancement; there is also a significant difference in perspective.
The respondent is more likely to look at each test conducted by the appellant
in isolation; the appellant takes a broader view of all the tests performed within
a project.
[10]
I note in passing the context in which the
appellant operates. The appellant performs, inter alia, sub-assembly operations
for a company that assembles dryers in Montréal. Considering the cost of labour
in Montréal, it is an operation that would have easily been relocated elsewhere,
such as Mexico or Asia.
[11]
The projects at issue are aimed at increasing
the efficiency of the appellant’s work obtained without excessive expense, which
the appellant could not support while still remaining competitive. Three of
those projects are related to the sub-assembly manufacturing of dryers; it is
precisely this search for efficiency which allows the appellant and its client
to avoid the offshoring of dryer manufacturing.
[12]
For the reasons that follow, the appeal will be
allowed.
The facts
[13]
Serge Caouette is the president and, at the time
of the hearing, the sole shareholder of Les Abeilles.
[14]
The company was founded in 1987 by Mr. Caouette’s
mother. In the early years, the company provided various packaging and printing
finishing services to tobacco companies. However, the restrictions on the
advertising of tobacco products forced the company to diversify its activities.
[15]
While continuing to perform printing finishing
activities, the appellant began providing packaging services for businesses
operating in the food and cosmetic industries.
[16]
The appellant also broadened its activities in
the industrial sector. That is when it started the assembly of mechanical
components.
[17]
In essence, its activities involve the assembly
of components supplied by the client. Once the sub-assembly is complete, it is
sent to the client who inserts it in its [Translation] “main line” to
obtain the end product.
[18]
The appellant began development projects in
2002. Its first application for scientific
research and experimental development dates back
to 2003. Since then, it has made several applications per year.
The projects concerned
[19]
For 2009, the appellant made applications to
obtain credit for scientific research and
experimental development for six projects. Only
two were approved. The other four were denied.
Projects related to the sub-assemblies for Mabe/General
Electric
[20]
Three of the four projects denied were related
to the sub-assemblies for Mabe. Mabe manufacturers General Electric dryers. These
three projects are as follows:
(a) 2007-01: New assembling process for motors for dryers.
(b) 2007-02: New assembling process for heating elements for
dryers.
(c) 2009-01: New assembling process for control panels using
pull flow with electronic sequencing.
[21]
The first two projects started in 2007 and were
accepted by the Canada Revenue Agency in the years prior to 2009.
[22]
The purpose of these projects is not only to increase
efficiency, but also to do so while meeting very stringent quality standards.
[23]
Mabe has a plant in Montréal which performs
assembly operations for General Electric. General Electric holds 49% of the shares
of Mabe and initiated the collaboration between Mabe and the appellant, as it
was General Electric that requested that certain items be produced at the appellant’s
facilities. However, Mabe and the appellant are the ones under contract.
[24]
It was in 2004 that discussions began with Mabe to
obtain contracts from Mabe and undertake the projects at issue. Mabe benefited
from collaborating with the appellant as the appellant’s projects were aimed at
increasing efficiency of production at the appellant’s and at Mabe’s plants.
[25]
Over the course of a project, extensive testing
is performed on the production line.
This testing typically lasts several hours.
[26]
When the appellant does not follow the normal process
of production, it is a [Translation] “deviation” from
normal production. There are three types of [Translation] “deviations”, according to the appellant:
(a) substitution;
(b) change in
engineering; and
(c) experimental
development.
[27]
The controversy here concerns only those tests that
represent the last type of deviation.
[28]
Prior to performing tests, approval from Mabe is
required.
An application for tests is made to Mabe and, if Mabe agrees, Mabe submits a
document entitled “DSI” providing authorization.
[29]
Mabe provides the components used for testing for
free and, after testing is completed, the sub‑assemblies are sent to Mabe
which sends them to Mexico to be disassembled.
[30]
Mr. Caouette explained that it was during
daily operational meetings that it was decided whether tests would be performed
the same day.
[31]
During testing, commercial production stops, changes
to be tested are installed or modified, the test is performed, everything is put
back together as it was before the test
and, then, production starts again. Mr. Caouette explained that when
performing a test, an attempt is made to ensure a large number of assemblies, thus
allowing an evaluation of quality and cycle time.
[32]
There are a series of specific problems that
must be resolved to achieve the objective sought and several tests could be
required to resolve one particular problem.
[33]
Each test relates to a change in process.
[34]
If the test is conclusive, or if the series of
tests is conclusive, the project will have to be approved by both the appellant
and General Electric before it can go to market. Once approved by the appellant,
the project goes to the first piece stage. This
involves, for example, producing approximately ten sub‑assemblies
and having them approved by General Electric. Then, the project must pass the pilot [run] A and pilot [run] B stages. Each stage
entails providing more parts to General Electric, which must then approve
production. During the pilot [run], the parts are
placed in the dryers and sold on the market.
[35]
However, the stages that took place after the
testing stage are not part of what is at issue here.
[36]
It is possible to retrace the steps of the
various projects through the log, the record or chronology of tests.
Mr. Caouette explained that the projects required many hours of work from
many individuals. At a minimum, it took as many employees to conduct a test as
it did to ensure regular commercial production.
[37]
There are also detailed descriptions of the projects
and tests produced.
Project 2007-01: motors
[38]
The appellant’s project had several objectives.
[39]
This project began in 2007 after the main conveyor
was received. The objective was to increase production efficiency, which was
measured in terms of the manufacturing time required, and to adapt the conveyor
so that it could assemble all families of motors.
[40]
Initially, the production of each different type
of motor at Mabe was carried out on a different carousel. There was a different
template for each type of motor. It took quite a long time to change the
templates on the carousel.
[41]
In early 2009, the conveyor was in commercial
operation and the appellant built one type of motor, regular motors. The
appellant still used the carousel (or carousels) for all other types of motors.
[42]
For 2009, the primary objective of this project was
to make it possible to assemble all the different types of motors on the same production line. It
was also hoped that a motor could be assembled every 9 seconds while meeting
the required quality standards of the client, that is, a maximum of 300 sub‑assemblies
of motors rejected per million sub‑assemblies, that is, a maximum of 3 rejects
out of 10,000 motors; at the beginning of the year, the appellant was at
approximately 14 seconds.
[43]
Not only did the appellant want to get rid of assembly
carousels, which were less efficient, but the appellant also wanted to be able
to change the production model without having to stop the production line so as
to be able to quickly change the model of motor at the client’s request.
[44]
For these objectives to be met, the following
problems had to be addressed:
(a) the adaptation of equipment to the various types of motors (pulley
press, jigs, mandrels);
(b) the synchronization of equipment.
[45]
On the line developed by the company, there is
only one motor template for all the families, rather than multiple templates on
the carousel. On the new line, no there are no parts to change when a decision
is made to produce a different model, which makes it possible to never stop the
line.
[46]
At the end of 2009, all families of motors could
be assembled on the line. The equipment was therefore adapted (pulley press,
jigs, mandrels) and the carousel was eliminated. The cycle time was reduced to
approximately 10 seconds.
[47]
Over the course of 2009, as part of this project,
32 tests were conducted. Appendix A to the judgment lists all the tests
and certain other work undertaken for the project.
[48]
In 2009, the appellant invested over 9,000 person‑hours
in this project. The breakdown of hours is provided in Appendix B to the judgment. The expenses claimed for this
project are approximately $137,000 in salaries and $1,300 in materials.
[49]
When the appellant began working with avec
Mabe/General Electric, they tried to find solutions already available to
achieve the required objectives.
[50]
The appellant’s agreement with the client provides
the appellant access to all knowledge available in the Mabe/General Electric
network, but despite all the experience available through the network, no one was
able to provide more than general principles; no one had specific solutions. The
appellant was unable to find any more information by speaking with its suppliers
or from Web searches.
[51]
The appellant itself had to find the necessary solutions
to meet the objectives.
Project 2007-02: heating elements
[52]
It was in 2005 that discussions with Mabe about
this project began. The idea was still to improve the productivity of Mabe’s
plant. Indeed, the assembly of the heating elements was done at the time on six
production lines. The appellant’s plan was to centralize this production on a
single line, while adhering to a cycle time of 7 seconds and a quality standard
of less than 300 rejects per million sub-assemblies.
[53]
In 2009, new models of heating elements had to
be added to the line. The required cycle time and desired quality standard
still had to be met
in a safe manner (hence the use of “foolproof devices”). To that end, it was
necessary, among other things, to address the following problems: too much
proximity between the heating element and the housing, false rejects caused by
the test station, instability caused by concrete dust, uncleaned lubricant deposit
formation in the equipment and overall stability.
[54]
At the end of 2009, the cycle time was 8.9 seconds
and the quality standard had yet to be met. However, the four families of elements
could be assembled on the same line and the proximity and false reject problems
were solved.
[55]
The proximity problem was solved by the use of a
cylinder system. This solution was the third of a series of three. The first was
manual and required a special tool. The second was hybrid and included a thumb
detector.
[56]
The false reject problem was caused by the impact
between the element and the stop designed to keep the element at the test
station. The solution found was to install a linear motion stop that did not
cause a shock that would move the element.
[57]
To Mr. Caouette’s knowledge, this type of assembly
line does not exist elsewhere. Indeed, neither the General Electric network nor
the appellant’s suppliers, the Web search or the manufacturer of the heating
elements were helpful in identifying a business that performed this type of assembly.
Moreover, General Electric and the appellant were the first in America to use the mica heating elements.
[58]
In 2009, the appellant conducted 19 tests and
invested over 7,500 person‑hours in this project. The appellant claimed
approximately $110,000 in salaries and $1,100 in materials.
Project 2009-01: control panels (backguards)
[59]
It was in 2009 that Mabe approached the
appellant to centralize the assembly of dryer control panels on a single line. The
objectives were to develop a system that (i) was capable of producing 174 variations
of panels, variations made on the basis of 600 possible components, (ii)
could achieve the particular variation ordered within four hours following the
order;
and (iii) could produce the panels much faster than Mabe.
[60]
To accomplish this, the appellant applied the [Translation] “surgeon approach” to its assembly method. This approach
consisted of one employee
who would put together the components necessary to assemble a given panel model
and bring everything to the assembler, who performs his or her work without
having to move around, like a surgeon who is provided with the tools required
for an operation.
[61]
The so-called [Translation]
“surgeon” approach was the appellant’s second
choice, as its first choice was too expensive to implement.
[62]
Not only was it necessary to validate the idea of
the [Translation]
“surgeon approach” applied to the assembly of
control panels, but it was also necessary to be able to assemble all models and
change from one model to the other within 30 seconds. Here again, the
quality standard sought was less than 300 rejects per million sub-assemblies
and foolproof devices were necessary.
[63]
In March 2009, it became apparent that it was
necessary to automate the communication process between the various
stakeholders, otherwise further improvements would have been impossible. To do
this, new software had to be developed as none of the existing software
products met the company’s needs. To that end, the appellant sought the assistance of subcontractor
ISG. The following modules were created: order automation module, import/export
module for transferring information between the company’s system and the client’s
system, survey module showing what the client has to produce, assembly module
ensuring that the assembler is provided with the right model, shipping module allowing
the client to confirm shipping of orders and know the status of the appellant’s
inventory of components.
[64]
The appellant did not claim any credits for the
“development” of the software, because ISG performed the task.
[65]
At the end of 2009, many objectives had yet to
be attained. The above-mentioned modules were nonetheless all developed and the
[Translation] “surgeon approach” was validated.
[66]
With regard to the problems encountered, a
problem of static likely to produce a spark causing the electronic controls to
burn out had to be resolved. The problem was solved with grounding placed
between the assemblers.
[67]
Assembly testing as such was not interrupted by
the automation of communication. Tests were still being conducted with respect
to the conveyor, templates, etc.
[68]
This project continued in 2010.
[69]
In 2009, the appellant conducted 22 tests and
invested over 9,000 person‑hours in this project. The appellant claimed
approximately $160,000 in salaries and $2,600 in materials.
[70]
Subsequently, Mabe/General Electric attempted to
apply the knowledge developed by the appellant in plants in Mexico.
Project 2009-02: printing finishings
(application of secondary components)
[71]
This is the only project with no connection to Mabe/General
Electric. Prior to 2009, the appellant had already established a printing finishing
line that made it possible to handle a variety of printing components
(magazines, cartons, advertising inserts) and to add secondary components; there
was a suction conveyor and various equipment (a primary feeder, a secondary
feeder, a glue applicator, a label maker, two folding modules, etc.).
[72]
These operations were all conducted at a nominal
rate of 6,000 applications per hour. Individually, the various equipment types
could operate much faster than together.
[73]
In 2009, the objectives were to (i) increase the
speed at which all operations could be conducted together to 11,000 applications
per hour; (ii) modify the secondary power supply so as to accommodate larger
components; and (iii) add new secondary components.
[74]
Among the issues to be addressed to achieve the key
objectives were the irregularity in the amount of glue applied to the components,
the positioning of a new module generating labels, the elements not detected by
the detection module and the synchronization of equipment.
[75]
In 2009, the secondary feeder that was modified to
accommodate larger components was installed and its stability was confirmed. By
the end of the year, the objective of 11,000 applications per hour still had
not been reached when all the modules were working together.
[76]
Once the testing was completed, the materials
were recycled, and not disassembled as in the case of the other projects.
[77]
To Mr. Caouette’s knowledge, a Chicago-based
company reportedly has similar facilities, but he did not wish to share that
knowledge. Mr. Caouette had not been able to find an assembly line already
capable of performing the required work. Certain modules, such as the glue
application module and the label application module, were purchased “as is” but
had to be adapted to the line developed by the appellant. Other modules were
entirely developed by the appellant, including the primary feeder, the detection
system, the secondary feeder, the modified secondary feeder and the folding modules.
[78]
In 2009, the appellant conducted 13 tests
and invested close to 6,000 person‑hours in this project. The appellant claimed approximately
$74,000 in salaries and $300 in materials.
Expert evidence
[79]
The appellant called Martin Gariépy as an expert
witness. Mr. Gariépy has a bachelor’s degree in pure mathematics, a
master’s degree in aerospace engineering and a doctoral degree in mechanical
engineering. He taught some courses at the École polytechnique de Montréal and carried
out various work related, inter alia, to aerodynamics.
[80]
Mr. Gariépy was recognized as an expert.
[81]
The respondent called Steven Kooi as an expert
witness. Mr. Kooi has a Bachelor of Science in chemical engineering and
master’s and doctoral degrees in mechanical engineering. Prior to working for
the Canada Revenue Agency, Mr. Kooi had 22 years of varied experience
in the industry.
[82]
He was a scientific advisor at the audit stage.
[83]
The appellant objected to Mr. Kooi’s recognition
as an expert witness. The appellant did not challenge Mr. Kooi’s training
and experience, but rather his independence. I took the objection under reserve
and allowed Mr. Kooi to testify. For reasons that will become apparent
below, it is not necessary for me to adress that objection. I note that what is
important is the impartiality of the expert witness rather than his independence.
[84]
The findings of the two expert witnesses are
that the projects are or are not, scientific research and experimental
development within the meaning of the Act. However, that is a question that
must be answered by the Court and cannot be the subject of an expert opinion. While the old rule that an
opinion is never admissible when it concerns the very question to be decided by
the judge has been discarded for some time now, “the closer an expert opinion comes to opining on the
ultimate issue in dispute, the
more the trial judge must scrutinize its probative value.”
[85]
I note that, generally speaking, it would have
been useful to have expert evidence that focused more specifically on the
current state of practices and knowledge respecting assembly methods and techniques.
[86]
I note that Mr. Gariépy’s report is relatively
general.
Mr. Kooi’s testimony
[87]
I have several difficulties with Mr. Kooi’s
testimony and report as an expert witness.
[88]
My first difficulty is the following. In his testimony
and in his report, there is some confusion between his role as a scientific
advisor during the audit and that as an expert witness.
[89]
As a scientific advisor at the audit stage, it
is completely normal that Mr. Kooi would be guided by the Canada Revenue Agency
guidelines with respect to scientific research and experimental development,
including certain proof of facts standards that the taxpayer is required to
establish to satisfy the Agency.
[90]
However, his role is different as an expert
witness, as it is his personal expertise on such matters as whether there is technological
uncertainty. An expert may agree with a recognized authority within a field, but
he or she must nevertheless form his or her own opinion.
[91]
In his testimony and in his report, there are
times when Mr. Kooi often seems to be guided more by the Canada Revenue
Agency’s guidelines and policies than his personal expertise.
[92]
For instance, Mr. Kooi gave considerable importance
to whether a degree of contemporaneous documentation exists as required by the
Agency.
At the audit stage, the Agency is at liberty to decide what the taxpayer should
normally do to convince it of certain facts.
[93]
However, in the course of an expert’s testimony,
he or she expresses an opinion on the basis of certain facts; it is not the
role of the expert witness to determine the facts. If there is a controversy about
the facts, it is for the court to decide what the facts are.
[94]
Whether contemporaneous documentation exists, or
not, and the fact that documents contain, or not, certain information are relevant
to the resolution by the Court of controversy about facts. However, the existence
of contemporaneous documentation, or contemporaneous documents with specific content,
is not a condition to the recognition of scientific research or experimental
development.
[95]
This confusion about roles is also illustrated
by a number of references to the requirements of the Canada Revenue Agency,
such as [Translation] “the Agency requires that the analysis take into account the
following probative evidence” preceding a list of 14 elements about
one page long.
[96]
The penultimate conclusion at page 47 of the report is as follows:
[Translation]
We conclude that
the documents adduced by Les Abeilles to support the testing claimed were not
contemporaneous with the testing. Despite the fact that materials used for the
testing were provided by the clients, and not claimed by Les Abeilles, the materials
used and applied in the testing are part of the supporting documentation
required for the progress, evolution and justification of the testing.
Once again, it
is a matter of determination of facts and not expertise.
[97]
I conclude from all of this that Mr. Kooi was
not impartial.
[98]
Before moving on to my second difficulty, I note
that Mr. Kooi’s emphasis on the absence of certain documents is such that
it is not always obvious what the factual basis of the opinion expressed is.
[99]
Second, in view of contain numerical errors, it
is quite clear that, in reaching his conclusions, Mr. Kooi’s perception of
part of the context was somewhat flawed.
[100] Specifically, it is clear that in preparing his report, the witness
believed that the amount of time claimed for production-line testing was
greater than it actually was.
[101] Mr. Kooi’s report regarding project 2007-02, reads, in part:
[Translation]
In reviewing Table
2, eighteen tests were conducted and between 11 (test 2456) and 56 persons (test
2282) were involved in said tests. We found that in these two extreme cases, the
estimated time spent on these tests and claimed as experimental production,
based on the machine operating twenty-four-seven, were 11 days and 26 days. The
justifications for these lengthy tests, such as test data, the report and reasons
for continuing testing following tests, were not established. It is difficult
for us to confirm the validity of the claim.
[102] It is apparent from this paragraph that the witness assumed that the
two production tests in question, 2456 and 2282, lasted the equivalent of 11 and
26 days, respectively, operating continuously on a 24-hour basis.
[103] In Table 2, there is a summary of all the tests conducted during the
year for project 2007-02. Test 2282 is the one with the most number of hours, 627.25 hours.
Test 2325 is the one with the least number of hours, 113 hours. That is
the equivalent of 4.7 periods of 24 hours, approximately five days.
[104] The total hours in Table 2 is over 6,000 [Translation] “hours of tests.” It is
the equivalent of 250 periods of 24 hours or 250 days.
[105] If that particular production line operated 365 days per year, 24 hours per day, the tests would represent over 68% of the
annual operation of the production line in question. However, if the line 24 hours per day, 5 days per week, it would be the equivalent of
close to 100% of the annual operation of the line.
[106] If that were the case, I could see how this would raise doubts from
a factual perspective given that it is an operational production line.
[107] This error does not occur only with project 2007-02. In the last
paragraph of page 43 pertaining to project 2009-02, the same type of presumption
of fact is made according to which the 537.5 hours represent a test that lasted
22 days.
[108] In the report, there is a Table 1,
similar to Table 2, pertaining to project 2007-01. The total [Translation]
“hours of tests” in Table 1 is over 8,000 hours or 330 periods
of 24 hours. In the light of how the witness construed the hours of testing, this
implies that he understood the claim as being for hours of tests, which is almost
the entire use of the production line for the year.
[109] However, in reviewing the evidence, it is clear that it is person-hours
of work over the course of testing, not hours of operation of the production
line. The tests lasted far less time than the witness thinks.
[110] In the paragraph cited above, the report also says that 56 persons
were involved in test 2282. However, the witness corrected that at the
beginning of his testimony and indicated that it should have been 28 persons
instead of 56.
[111] More generally in Table 2, he corrected the column entitled [Translation]
“Number of persons in Table 2;” the number indicated on each line must, as a
general rule, be divided by two.
[112] I conclude that, in preparing his report, Mr. Kooi’s findings
were drawn from, inter alia, a factual basis where the tests pertaining
to the four projects in question accounted for the primary use of the four production
lines for the majority of the year. The evidence shows that the tests lasted
far less time.
[113] Such a contextual error must necessarily affect one’s opinion.
[114] For these reasons, I give very little weight to Mr. Kooi’s
testimony as an expert witness.
[115] However, insofar as Mr. Kooi testified about what he did as a
scientific advisor at the audit stage, I accept his testimony, but I note that
it is not expert evidence.
[116] The reasons that led Mr. Kooi, as a scientific advisor, to
conclude that it was not experimental development are summarized in his technical review report dated November 18,
2010.
[117] Except for project 2009-01, Mr. Kooi concluded that there was
no technological obstacle, as they were engineering challenges and the solutions
were based on current standard engineering practice.
[118] For example, his conclusion regarding project 2007-01 is as follows:
[Translation]
. . . we found
that tests were conducted to address issues that did not constitute a technological
obstacle. The issues to be addressed for these projects as described were as
follows. . .: improvements in the hub of the pulley press capable of working on
all types of motors, the alignment of motors, the alignment of pulleys,
improvements in cycle time, the validation and confirmation of the operation of
equipment, etc., are engineering challenges. The solutions applied to address
these issues are based on current standard engineering practice by trial and error.
We conclude that, with respect to this claim, some of the work performed is
related to the application of developed technology to a new situation to
stabilize the process and improvement of several assembly stations.
[119] The case of project 2009-01 is a little different, as the report
concluded that [Translation] “the work was uncorroborated” and therefore no opinion was
provided.
[120] In reading the report dated November 18, 2010, given the importance Mr. Kooi
attached to it during his testimony, there is surprisingly no mention of a lack
of contemporaneous documents.
A note on the facts and documents
[121] It is useful at this point to note that the documents were filed by
consent and their content was not disputed.
[122] Therefore, I assume that the factual descriptions of the projects
and tests in these documents correctly reflect the purpose and accomplishments
of the projects as well as what was done during testing.
Mr. Gariépy’s testimony
[123] Mr. Gariépy began testifying by providing some definitions.
First, he explained what “scientific
uncertainty” means. On the one hand, he cited the definition by the
Canada Revenue Agency and, on the other hand, he gave his interpretation of the
phrase. According to him, there is scientific uncertainty
when a specific objective is identified but it is unknown whether and how it
will be achieved. He then defined the concept of “systematic
investigation.” This means that once the
uncertainties have been identified, a literature review will have to be
completed to find existing solutions, make hypotheses and perform tests to
support or disprove them. He added that the systematic investigation
need not assume a particular form.
[124] Finally, he explained what “technological
advancement” is. It is, based on his understanding
of the Agency’s policy, an advancement of the company’s knowledge or processes
that is not easily accessible; for example, such an advancement is not
attainable through a literature review or the purchase of a machine.
[125] Then, Mr. Gariépy provided his opinion on the presence of
uncertainties in the various projects. With respect to project 2007-01,
pertaining to the assembly of motors, Mr. Gariépy was of the view that it
was obvious that there were scientific uncertainties
in this project. It was unknown, at first, whether the objective of achieving a
nine-second cycle time was attainable. The same is true for the basic objective
of assembling all motor types on the same line.
[126] As for project 2007-02, regarding heating elements, there were
uncertainties with respect to cycle time, rejection rates and the assembly of
the various models on the same line.
[127] As for project 2009-01, pertaining to control panels (backguards),
Mr. Gariépy stated that the sum total of what was being sought was
uncertain. Although lean manufacturing is a known
theoretical concept, its practical application requires development.
[128] Finally, project 2009-02, regarding printing finishings, was also
uncertain. The fact that each separate element has the potential of achieving
the intended objectives does not preclude uncertainty about the overall objective
of having everything work together.
[129] Mr. Gariépy stressed the fact that it is necessary to look at
the projects [Translation] “from the highest level,” that is to say, that it is necessary to look at a project as a
whole. To look at every little step would be tantamount to distorting the
project. He noted that it is possible that a secondary objective does not
represent uncertainty, which does not preclude the validation of the project as
a whole. Later in his testimony, Mr. Gariépy stated that, according to his
interpretation of the Canada
Revenue Agency’s directives, it is necessary to look at
the project in its entirety, not just in the fiscal year concerned.
[130] According to Mr. Gariépy, there is no question that there was a
systematic investigation on the
appellant’s part. Team meetings were held to determine objectives, accessible
solutions were sought in the literature, tests were planned and documented, and
conclusions were drawn.
[131] Mr. Gariépy then provided his opinion about technological
advancement in each project.
For project 2007-01, (the fact of )attaining 10.9 seconds is an advancement. The
adaptation of the hub, of the pulley press, etc., are advancements.
[132] As for project 2007-02, numerous advancements were also made. The
mere fact of making the assembly process more stable is an advancement.
[133] Project 2009-01 itself is an advancement. A [Translation] “new assembly line concept” was developed that
did not exist before, at least not in practice. In Mr. Gariépy’s
view, a major indicator of the level of advancement was the subsequent
technological transfer to Mexico.
[134] As regards project 2009-02, the more than 50% increase in the number
of applications per hour by increasing synchronization is a major advancement.
[135] Mr. Gariépy was critical of the report by Mr. Kooi, the
respondent’s expert, on the grounds that he looked at the projects far too
closely, instead of looking at them from the highest
level. He also expressed doubt as to Mr. Kooi’s understanding of
the various projects.
[136] Mr. Gariépy then addressed the issue of Translation] “current practices” that is often
referred to in Mr. Kooi’s report. According to Mr. Gariépy, [Translation] “current practices are processes or methods. .
.that are introduced on a regular basis.” Again,
his definition is a clear and simple explanation of the Canada Revenue Agency’s
definition. According to him, there is no harm in resorting to current practice
if it can help us achieve our objective.
[137] In his cross-examination, Mr. Gariépy explained that the fact
of completing a project at a lower cost when the technology is available
elsewhere, but at a greater cost, can represent an advancement. He then
referred to the software developed as part of project 2009-01.
Analysis
[138] The central issue is: Do the four projects in question constitute “experimental development” as defined in the Act?
[139] The phrase “experimental development” is
defined as follows:
. . . work undertaken for the purpose of
achieving technological advancement for the purpose of creating new . . .
devices . . . or processes, including incremental improvements thereto,
[140] Thus, the following questions must be answered:
(a) Were the projects undertaken for the purpose of achieving
technological advancement?
(b) Were they
undertaken for the purpose of creating new processes, including incremental
improvements?
[141] The case law has developed a number of useful criteria to determine
whether or not activities constituted scientific
research or experimental development. These
criteria, which were listed by Judge Bowman, as he then was, were approved by
the Federal Court of Appeal. These criteria are summarized in CW Agencies
Inc. v. Canada,
where the Court of Appeal states as follows:
17 Both sides in
front of us relied on the test outlined in Northwest Hydraulic Consultants
Limited v. Her Majesty the Queen, 98 D.T.C. 1839. In that case, Judge Bowman of the Tax Court outlined five
criteria which are useful in determining whether a particular activity
constitutes SRED. Those criteria have been approved by this Court in RIS-Christie
v. Her Majesty the Queen, 99 D.T.C. 5087 at page 5089. The criteria are as follows:
1. Was there a technological risk or uncertainty which could
not be removed by routine engineering or standard procedures?
2. Did the person claiming to be doing SRED formulate
hypotheses specifically aimed at reducing or eliminating that technological
uncertainty?
3. Did the procedure adopted accord with the total discipline
of the scientific method including the formulation testing and modification of
hypotheses?
4. Did the process result in a technological advancement?
5. Was a detailed record of the hypotheses tested, and results
kept as the work progressed?
[142] It must be borne in mind that these crateria are used to help
determine whether or not a
technological advancement has occurred. The first crateria,
technological uncertainty, is one way of dealing with the technological advancement criteria; there can hardly be a technological advancement if one
already knows how to achieve the end result; the second and third criteria are,
inter alia, one way of ensuring that the work was undertaken for the
purpose of achieving technological advancement and that it was not, for
example, an advancement achieved by accident rather than work undertaken for
the purpose of achieving technological advancement.
[143] The five criteria are not absolute. For example, there is no
requirement that the work must result
in a technological advancement; if the work was
unsuccessful but undertaken for the purpose of achieving technological
advancement, it may still qualify.
[144] I will begin with a review of project 2007-01, motors for dryers.
[145] The evidence is very clear.
[146] There were two objectives at the beginning of the year: first, to
increase the production rate for the sub‑assembly of motors from every
14 seconds to every 9 seconds. Second, the appellant sought to produce all
the different types of motors on the same production line, which, at the
beginning of the year, could only produce [Translation] “regular” motors without having to stop production to change the motor size.
[147] At the end of the year, the appellant reached a production
sub-assembly of approximately every ten seconds, an increase of about 40%
and the appellant was able to carry out the sub-assembly of all motor sizes on
the same line. As a result, not only was there an increase in productivity, but
there was also a dramatic increase in the flexibility of production as the
appellant was able to change the size of the motors many times per day.
[148] There is no doubt that not only did the appellant seek significant
improvements in its production processes, but there were also significant
improvements in the production processes of motors.
[149] Is this technological
advancement, or the application of standard procedures
or routine engineering, as the respondent contends?
[150] It is true that when reviewing individual tests they often do not
appear to be, in and of themselves, a significant advancement.
[151] In argument, the respondent took the position that Mr. Gariépy
went too far by looking at project 2007-01 in its entirety from its outset in
2007.
[152] I fully agree that experimental development was required in the year
in question. However, this does not mean that one cannot examine the history of
a project that began in a previous year by considering whether, in the
particular year at issue, “experimental
development” was carried out within the meaning
of the Act.
[153] Furthermore, it is important to consider each project globally in
the year and not each test individually.
[154] The evidence is very clear that the appellant did not know at first
how it would go about increasing the production rate of motors from every 14
seconds to every 9 seconds; nor did it know how it would go about changing the
production line to able to produce all motors on the same line.
[155] The appellant had access to the knowledge network of Mabe/General
Electric. Mabe/General Electric engineers were only able to provide general
principles. The appellant was unable to find a ready-made solution by speaking
with its suppliers or by conducting Web searches.
[156] The appellant had to come up with its own solutions, at relatively
modest cost—its total claim for this project was less than $239,000.
[157] I accept, as Mr. Gariépy testified, that there was “systematic investigation”. This can be seen very
clearly in the numerous documents produced, including those found in Tabs 8 and
9 of Exhibit I-2. It is clear from these documents that hypotheses were
formulated and that the scientific method was
applied.
I note that this is also true for the three other projects.
[158] To find solutions, the appellant had to conduct 32 tests and
invest over 9,000 person-hours. For each test the appellant had to make
necessary changes to the production line and, after the testing, the appellant
restored the line to the same state it was in prior to the changes made for the
tests. The sub-assemblies were then disassembled.
[159] These were not tests where adjustments were made to a known
production process. They entailed a series of changes, some of which helped
with the objectives and others that were unsuccessful. There was no certainty
of the result, and indeed, the appellant, despite all of its efforts, was
unable to achieve its sub‑assembly objective of every nine seconds in
2009.
[160] As for the tests that were successful, it was only later, with the
approval of Mabe/General Electric, that the changes were finally put in place
on the production line.
[161] I do not see how, under these circumstances, what the appellant did
in 2009 can be classified as an application of standard procedures or routine engineering.
There was technological uncertainty.
These were not standard procedures.
[162] At the end of the year, the appellant had a new process that
consisted of certain equipment arranged and adjusted in such a specific way so
as to henceforth produce more quickly and with greater flexibility.
[163] There is no doubt that it is technological advancement and that the
work was undertaken for that purpose.
[164] The situation is rather similar for the other projects.
[165] In the case of project 2007-02, in early 2009, the appellant had
three objectives: first, to achieve the speed sought—sub‑assembly
every 7 seconds—but at the beginning of the year it had only achieved
9.6 seconds; second, to achieve the desired quality standard, less than
300 rejects per million; third, to integrate new models of heating
elements in addition to the [Translation] “regular” model
that was already in production.
[166] Again, the appellant did not know how it was going to achieve its
goals. Information on how to achieve them did not exist elsewhere.
[167] As in the first project, the appellant had to conduct numerous tests
to find solutions. Specifically, the appellant conducted 19 tests and
invested over 7,500 person‑hours in this project. As was the case
for the other projects, most of the tests consisted of a significant number of
production hours and a significant amount of production. The sub-assemblies
produced were disassembled in Mexico as were those produced for the first
project.
[168] As part of the project, the appellant undertook a systematic
investigation and, at the end of the year, it was partially successful in
achieving its objectives. It successfully integrated the new models of heating
elements, but it was only partially successful in achieving its objective in
relation to assembly time. The appellant achieved a time of 8.9 seconds,
an improvement of slightly less than 8%.
Finally, as for the quality, the goal had yet to be attained as of the end of
the year.
[169] This came at a rather modest cost of approximately $111,000.
[170] Again, I have no doubt that there was technological uncertainty and
that it was experimental development within the meaning of the Act.
[171] As for project 2009-01, regarding dryer control panels, it was a
very ambitious project that began in 2009. With the [Translation] “surgeon approach” the
appellant wanted to be able to have very flexible production processes while
improving productivity when compared to what it originally was at the
Mabe/General Electric plant.
[172] It is not surprising that this project had significantly more
difficulties when one considers how ambitious the objectives were: a system
(i) capable of making 174 variations of panels; (ii) ensuring the
attainment of the particular variation ordered within four hours of the order
being placed; and (iii) while increasing the production rate.
[173] The appellant systematically carried out the project. In late 2009,
the appellant validated the [Translation]
“surgeon approach,” but overall it was not
successful in achieving its objectives, despite 22 tests, approximately
9,000 person‑hours and $163,000 invested in the project.
[174] It is clear that the appellant did not know how it was going to
achieve its objectives and that there was technological uncertainty. It was
work undertaken for the purpose of achieving technological advancement. The
procedures were not standard.
[175] Finally, as for project 2009-02, regarding printing finishings,
early in the year the appellant’s objective was to increase by over 80% the
number of secondary components that could be added to printed materials, enable the secondary feeder
to accommodate larger components and add new secondary components.
[176] The appellant conducted 13 tests and invested over
6,000 person‑hours and approximately $74,000.
[177] Again, it is clear that there was technological uncertainty; the
appellant did not know at the outset how it was going to achieve its goals and,
indeed, it was only partially successful in achieving them—at the end of the
year the secondary feeder could accommodate larger components, but the goal of
11,000 applications per hour had yet to be attained. The appellant tried
to obtain the necessary knowledge elsewhere, but was unable to do so and therefore had to develop
its own solutions.
Conclusion
[178] In summary, the four projects in question constitute experimental
development within the meaning of the Act.
[179] Accordingly, the appeal will be allowed, with
costs, and the matter is referred back to the Minister of National Revenue for
reconsideration and reassessment on the basis that projects 2007-01, 2007-02,
2009-01 and 2009-02 constitute experimental development within
the meaning of the Act.
Signed at Ottawa, Ontario, this 23rd day of October 2014.
“Gaston Jorré”
on this 15th day
of June 2015
François Brunet, Revisor
Email this Content