In this research, a novel algorithm for real-time orbit determination (RTOD) is presented using the robust unscented Kalman filter (RUKF) with global positioning system (GPS) group and phase ionospheric correction (GRAPHIC) observables. To increase the reliability of the solution, a robust approach is included in the UKF to cope with the bad, invalid, or degraded measurements leading to the divergence or inaccurate output of the filter. Robustness is provided by making the filter less sensitive to faulty measurements using a scale matrix that is multiplied with the covariance matrix of the observation noises. Real data collected during a massive solar storm are used in the algorithm. For external validation, the outputs of RUKF and classical UKF are compared with the precision orbit ephemerides of the Challenging Minisatellite Payload (CHAMP). The results show that RUKF slightly outperforms classical UKF and possesses the capability to be used as an efficient and reliable algorithm in case of bad observations or malfunctioning of the system. (C) 2017 American Society of Civil Engineers.