We present 1.3 mm continuum ALMA long-baseline observations at 3-5 au resolution of 10 of the brightest discs from the Ophiuchus DIsc Survey Employing ALMA (ODISEA) project. We identify a total of 26 narrow rings and gaps distributed in 8 sources and 3 discs with small dust cavities (r $<$10 au). We find that two discs around embedded protostars lack the clear gaps and rings that are ubiquitous in more evolved sources with Class II SEDs. Our sample includes 5 objects with previously known large dust cavities (r $>$20 au). We find that the 1.3 mm radial profiles of these objects are in good agreement with those produced by numerical simulations of dust evolution and planet-disc interactions, which predict the accumulation of mm-sized grains at the edges of planet-induced cavities. Our long-baseline observations resulted in the largest sample of discs observed at $sim$3-5 au resolution in any given star-forming region (15 objects when combined with Ophiuchus objects in the DSHARP Large Program) and allow for a demographic study of the brightest $sim5%$ of the discs in Ophiuchus (i.e. the most likely formation sites of giant planets in the cloud). We use this unique sample to propose an evolutionary sequence and discuss a scenario in which the substructures observed in massive protoplanetary discs are mainly the result of planet formation and dust evolution. If this scenario is correct, the detailed study of disc substructures might provide a window to investigate a population of planets that remains mostly undetectable by other techniques.