The first of our Three Tests for a project is novelty - we want to progress the state of the art and do something new in our field.

But how do we define novelty?

We look at it from a couple of viewpoints: what we define as novel within Fuzzy Labs; and what external bodies - especially funding bodies and UK tax rules - define as novel. Given we want to be able to apply for funding to help with advancing the state of the art, the latter is important!

To be eligible as a Fuzzy Labs² project, at least one of the following definitions needs to apply, although the strongest projects will meet multiple criteria.

Fuzzy Labs definitions [WIP]

MLOps novelty

A new tool, process or approach for MLOps - something that we can’t find evidence of existing already

Progressing state of the art

More than just novel or useful to us - something that is beneficial to the wider community and may be the subject of a research paper

Accelerating adoption

Taking an aspect of MLOps theory that may already have been the subject of research elsewhere, but has not been operationalised and is not widely known or adopted in the industry

HMRC definitions

R&D tax relief is only given if the government’s definition of R&D is met. This gives us a clear steer on a definition for novelty - if it meets the definition for R&D, it is a good indicator of novelty and importantly allows us seek financial support for our innovations in Fuzzy Labs²

In summary, to qualify as R&D, our projects must demonstrate all of the following:

• looked for an advance in the field
• had to overcome the scientific or technological uncertainty
• tried to overcome the scientific or technological uncertainty
• could not be easily worked out by a professional in the field

<aside> 💡

We will use both the Fuzzy Labs and HMRC definitions above as tests of novelty for our project research questions

</aside>