Re-running a DICE

[ankostis]

We are opening this issue for public discussion on when it is allowed to re-run a DICE, and how. The two reasons for doing that are:

1. to extend the Interpolation Family to Higher/Lower vehicles, and
2. to correct WLTP data (corrupted or mistaken) or NEDC's CO2 declared-values.

(please reply below if you see is another reason)

In general we suggest to avoid re-dicing, because on every dice, the sampling decision must be respected (possibly re-testing the vehicle in WLTP/NEDC), and consequently, the verification-factors must be updated.

So far CLIMA has approved a single re-certification, based on reason no-1. That was a clear case involving a new physical WLTP test. For that, JRC implemented the --recertify feature of CO2MPAS in v1.7.3post0, which preserves the link between both runs, but only the last one is considered "ofrficial". Of course, re-dicing after the vehicle has been put in circulation would complicate things.

We have also forward requests regarding no-2 to CLIMA (errors in the inputs, with or without performing new physical WLTP-tests), and we will explain here the general outcome (criteria, rules applied).