There are many possible candidates for R&D KPIs, but most of them will make sense only on a big scale. To focus the efforts I’d suggest to use the steps of KPI system.
For example, in your very case what are the R&D process bottlenecks? Is the main challenge related to the collection of the ideas or prototyping? Once the focus area is defined you can track your efforts on time/cost/income/number scale.