Lightning research needs on-demand lightning strikes, because of the random character of natural lightning. Lasers have been proposed as alternatives to the current technique using rocket-pulled wires, because they would expectedly provide more flexibility. However, highenergy, nanosecond lasers cannot provide long connected plasma channels. In contrast, we recently reported the triggering of electric events in thunderclouds using ultrashort laser pulses. Further improvements of the laser pulse sequence and experimental geometry are discussed.