In this set of articles, we will elaborate (and give concrete examples) on the questions of the R&D Advance Assurance.

One of the questions you will be asked to answer is: 'Why is the knowledge being sought not readily deducible by a competent professional?'

Note that a competent professional or expert in the industry is someone in your company who has extensive knowledge of your industry. This expertise could be acquired through working in the industry previously or by studying the subjects relevant to the field at University. This person must have knowledge of existing solutions, be able to assess which activities would be considered an advance in science and technology and differentiate them from routine activities. Hence, they should be able to identify which activities would be considered R&D.

To answer the question about why the knowledge being sought is not readily deducible by a competent professional, provide the reason why the answer to the problem is not known using the current state of knowledge.

Note: Scientific and technological activity that involves a review of the field by a qualified professional, or which involves applying known techniques to find the answer to a problem, is not R&D. Nor does R&D occur where the knowledge is new to the company but not new to the wider world.

Example of an answer to the question: 'Why is the knowledge being sought not readily deducible by a competent professional?'

One of the main components in BAGE is its’ unique conversational agent, responsible for processing the incoming questions from users and for outputting their answers. The developers are creating a way to semantically represent natural languages (like English) into a machine readable format, which is hard to implement. The difficulty comes from the fact that machines understand strict, unambiguous, formal logical statements, but in natural languages semantical ambiguities appear very often. This means that even when the syntax and the meanings of the individual words in a sentence have been resolved, there might be more than one way of interpreting the meaning of the sentence. Language comprehension is an easy task for the human brain, since we can deduct the meaning from the context, but is rather more difficult for a computer and its’ solution requires complex algorithms to interpret the sentence in its’ intended way. Hence, the team of competent professionals are not certain if this would be feasible to achieve.

The team, along with our experts in the field spent many hours researching existing

techniques which aim to deal with this difficulty, but concluded that none of them could be readily applied to their project. They determined that the problem could be solved by splitting the conversational agent into three main sub-systems and working extensively on their functionality.

The first one, the natural language understanding module, parses the input using

customised functions, developed by the team. The module also includes a non-verbal

input system, especially useful for children. Many of the existing systems, mainly those

implementing speech-to-text solutions, are not reliable when handling children’s input – issues including pronunciation, accent, and timing are very common to youngsters and the existing technologies struggle to interpret their input accurately. B Games thought of innovative solution in the form of graphical user interface, which will allow children to construct their question. On the screen, children will also see alternative words and question paths, which will aim to inspire them to ask more questions.

Need more help? Any further question on your R&D Advance Assurance - please hit the chat button!

And if you haven't talked with us prior to starting your R&D Advance Assurance, feel free to book a call with one of our experts here!

Did this answer your question?