There is a distinction between experimental development and routine development. Experimental development activities are undertaken to resolve a technical uncertainty to meet a desired technical objective. In this case, the technical uncertainty manifests itself as an obstacle that requires a systematic investigation to resolve. When the technical objective can be achieved through routine practice then development activities are executed to achieve the technical objective. In this case, we often say there is a direct line of sight to the technical objective. As such, routine development is distinguished from experimental development when a direct line of sight to the technical objective is within the knowledge base of the organization to achieve.

Within the definition of SRED is the notion of a systematic investigation or search. Within this definition, the terms systematic investigation or search are mutually exclusive. As such, the scope of an investigation allowed by this definition is broad. When an investigation is systematic it implies there is an experiment that generates data that is then analyzed. In those circumstances where an investigation is conducted by means of a search it implies the phenomenon studied is through the analysis of existing data.

We have broadly discussed what constitutes a systematic investigation and how it is differentiated from a search. A systematic investigation requires the thoughtful design of an experiment followed by the analysis of data that is the consequence of an experiment. Such an experiment can validate a deterministic model that advances our scientific knowledge with or without a practical application in mind. Alternatively, statistical methods may be employed that use probabilistic models that explain a physical system and advances our understanding of both science and/or technology. As such, a systematic investigation requires both an experimental plan and the analysis of data that is derived from the execution of the experimental plan, whereas a search does not require an experimental plan. Often, a search may involve the re-examination of data that has already been collected and requires analysis to render new information. Such is the case in the medical field when multiple investigations are brought together. In this case, a meta-analysis of the data seeks to address new research questions. Therefore, a search is consummate with analysis.

A systematic investigation may involve experiments designed to confirm deterministic modelsthat describe physical systems. Deterministic models are mathematical expressions that precisely determine outcomes through known scientific relations without any room for random variation. In such models, a given input will always produce the same output. Alternatively, astatistical model is a mathematical expression which describes the behavior of a physical system in terms of random variables and their associated probability distributions.

Engineers and scientists try to advance scientific knowledge through the discovery of new deterministic models to explain real phenomena. For example, the Ideal Gas Law is a deterministic model based on the principals of science and confirmed through experiments designed to test if the deterministic model sufficiently predicts the physical system under investigation. While the Ideal Gas Law has proven useful over the years; we must recognize that no ideal gas exists. As such, even under the best experimental conditions, all experimental data is fraught with error and cannot always be controlled. The degree of agreement with a deterministic model may suffer and not sufficiently explain or predict the physical system under investigation and may require alternative approaches to experimentation and analysis based on statistical methods.

Statistical methods provide a structured and systematic approach whereby engineers and scientists can derive probabilistic models that explain a physical system under investigation. Using statistical methods to conduct experiments differs from experiments designed to confirm deterministic models. Statistical methods provide researchers and engineers an approach to experimentation and analysis that does not depend on whether or not a deterministic model and its scientific relations are known. Generally, statistical methods advance our knowledge through a series of experimental activities in which we make conjectures about the physical system under investigation, perform experiments to generate data, and then use this information to establish new conjectures that lead to new experiments, and so on.

When the data are subject to experimental error; the use of statistical methods is the only objective approach to experimentation and analysis. They lead to new insights that may assist the engineer or researcher in develop new deterministic models and scientific relations or reveal existing ones that explain the physical system under investigation. Statistical methods have found broad application in many disciplines of science and engineering and are responsible for much technological and scientific advancement.

In addition to the definitions of SR&ED, the income tax act also defines the types of work that support SRED as;

• Work with respect to engineering, design, operations research, mathematical analysis, computer programming, data collection, testing, and psychological research, where such work is commensurate with the needs, and directly in support of basic research, applied research, or experimental development.

This definition is broad and encompasses the many types of work that may be carried out within an academic or industrial setting. However, such work must be in direct support of SRED to qualify for an income tax refund or credit as opposed to work that is routine and not classified as basic research, applied research or experimental development.

Routine work is typically associated with work tasks required to maintain commercial operations. Note the difference between routine work and scientific research or experimental development stems from the fact that routine work does not try to increase scientific knowledge as is the case in basic research or applied research or achieve a technological advancement as is the case in experimental development. However, when routine work is engaged to support efforts to increase scientific knowledge or achieve a technological advancement then that work may qualify as a SRED activity.

The Income Tax Act also states that SR&ED does not include the following work:

• market research or sales promotion;
• quality control or routine testing of materials, devices, products, or processes;
• research in social sciences or the humanities;
• prospecting, exploring, or drilling for or producing minerals, petroleum, or natural gas;
• the commercial production of new or improved material, device, product, or the commercial use of a new or improved process;
• styles changes
• routine data collection.

Again, the reason that such work itself does not qualify is the fact that it is routine and does not increase our understanding of technology or scientific relations.

Within the definition of SRED there are two broad fields where a systematic investigation or search may be carried out, they are; science and technology. Within each of these two broad fields, there are three criteria to be met for any body of work to be considered as Scientific Research or Experimental Development. The following table lists the criteria for both Scientific Research and Experimental Development.

Common to both Scientific Research and Experiment Development is the nature of the advancement sought; the uncertainty in being able to achieve such advancement and the documentation or content that demonstrates the work that was performed.

The work that is performed through a systematic investigation or search must generate information that creates "new knowledge" thereby advancing our understanding versus the area being investigated. In the case of Scientific Research, it must advance our understanding of scientific relations versus experimental development where the advance is in the understanding of the technology under investigation. Advancing our understanding of scientific relations refers to evolving our understanding of deterministic models that describe physical phenomena. Experimental development, on the other hand, often uses statistical methods to advance our understanding of technology.

Within the context of the taxpayer (business) when a new or improved material, device, product, or process is realized it must embody either a scientific or technological advance to be eligible for a SRED tax credit or refund. In most cases, the taxpayer (business) is interested inexperimental development activities that resolve technical obstacles and yield advancements in technology that support business objectives. The degree of technological advancement is often based on the advancement in the capability or performance of the physical system. It can be a feature or capability which did not exist in the past or an enhancement in new or existing materials, devices, processes or products that enhances standard performance.

The presence of technical uncertainty (technical obstacle) is the differentiating factor between routine development and experimental development. When we experience a technical obstacle, it can be classified into one of three buckets. If a clear path exists to resolve a technical obstacle, then all that is needed is routine development or engineering to achieve a resolution - we call this a Direct Line of Sight solution. In those cases when more than one route reveals itself and it is uncertain which route is best then experimental development is required to remove such uncertainty and identify the best route. Finally, there are situations when there are no apparent routes. As such, it is uncertain if the technical uncertainty can be resolved at all. In such circumstances Scientific Research or Experimental Development will be incremental and follow a number of investigations that grows our scientific or technological understanding incrementally by breaking down the nature of the technical problem into manageable levels of technical uncertainty.

Both Scientific Research and Experimental Development activities must follow a systematic investigation or search. There must be the formulation of a hypothesis, which is evaluated through experiment and/or analysis followed by a conclusion. The documentation of this progression forms the scientific or technical content required to justify a SRED claim. Along with this criterion, there is a requirement that qualified personnel shall direct or perform the scientific or technical work or have the relevant experience in the appropriate field of science, technology, or engineering.

What is Scientific Research and Experimental Development

The Canadian Income Tax Act defines scientific research and experimental development, also referred to as SR&ED or SRED as a;

"systematic investigation or search carried out in a field of science or technology through experiment or analysis."

First and foremost we must make a distinction between scientific research and experimental development. The Income Tax Act defines scientific research (SR) within the context of basic and applied research in the following manner:

1.  Basic research, namely, work undertaken to advance scientific knowledge without a practical application in view;
2. Applied research, namely, work undertaken to advance scientific knowledge with a practical application in view.

Whereas experimental development (ED) is described as;

1. Work undertaken to achieve technological advances for the purpose of creating new or improving existing, materials, devices, products, or processes, including incremental improvements thereto.

Given the distinctions between Scientific Research and Experimental Development (SR&ED or SRED) mentioned, it is ease to see that basic research is fundamentally conducted within an academic setting. Applied research is often a collaboration between academia and industry where such a collaboration yields a tangible solution that addresses a market need having economic value to the organization. Experimental development, on the other hand, is often specific to the organization (taxpayer) and initiated when market conditions or commercial realities present technical obstacles that must be overcome. When organizations conduct experimental development, the product of such work usually results in new or improved technologies that support the growth and/or efficient operation of the organization. When the requisite knowledge is beyond the reach of the organization, then collaborations with an academic institution may be warranted. Such is the case in applied research when the nature of the investigation requires an advance in scientific knowledge to support a practical application.