PPP is a relatively new but powerful technique for positioning. The main difference between PPP and standard positioning is the use of carrier-phase measurements, whose noise is lower by two orders of magnitude with respect to code measurements. It is now widely accepted that PPP techniques can achieve centimeter-level accuracy globally in real-time, in particular when they are combined with phase integer ambiguity resolution. However, one important drawback of PPP is the convergence time. Dual-frequency PPP convergence is long, tens of minutes, which makes it impracticable for many applications. Until now, some improvement was made using a third frequency (L1/L2/L5 for GPS, E1/E5a/ E5b for Galileo), showing that quasi-instantaneous convergence was possible. In July 2018, a collaboration between NRCan and CNES showed that instantaneous (epoch-by-epoch) convergence is possible thanks to the use of the new E6 Galileo signal. 15 cm accuracy is achievable using Galileo alone and centimeter-level accuracy can be obtained by combining Galileo and GPS measurements (GPS World, July 2018 issue, Innovation column). In this poster, we present the real-time implementation of this new method, called OEUFS (for Optimal Estimation using Uncombined Four-frequency Signals), in the PPP-Wizard demonstrator. We first detail the method to compute the E6 phase bias in the demonstrator. This is challenging because the Galileo constellation is not fully deployed yet and the number of compatible stations is reduced. Thus, the E6 bias is not available everywhere at all times, but the validity of the method can still be assessed. We characterize this bias in terms of noise and time stability, and show that it can be casted into the SSR representation proposed by the RTCM for phase-bias messages. Then, we detail the user level implementation. Since OEUFS involves optimal partial ambiguity resolution, the BIE (Best Integer Equivariant) method has been coded and tested. Finally, some actual real time positioning results are presented.