Considerable efforts have been made during the last decades to develop HT phenotyping platforms, resulting in massive numbers of images and image sequences of plant root systems. However, the information content of these images is largely underexploited as available image analysis tools are restricted to a handful of morphological features and AI depends on the availability of accurate ground truth information. In this presentation, we show how focusing our attention on root tips, which are the generators of root architecture, opens alternative routes to unlock architectural information from root system image series. The extraction of time series of root tip features is a routine and quick operation, and the resulting dataset can be mined in different ways. Methods inspired from tip tracking can be used to reconstruct vectorized root systems with temporal information on individual segments. Tips can also be considered as samples that can be analysed with spatial statistical methods or used to parameterize density or structural root architecture models. We illustrate the latter route with a collection of 150K images of 500 wheat genotypes, acquired during HT phenotyping experiments. From a dataset containing the time series of tip coordinates, we have estimated in R several key parameters of two root architectural models, ArchiSimple and RootTyp. We found significant genotypic differences for several of these parameters, even though broad-sense heritabilities tends to be low. Finally, shifting our focus from an image analysis problem to a data mining one was also an opportunity to get closer to our scientificobject.