The fragmentation after blasting is crucial for the efficiency of the operation in the sublevel caving mine in Kiruna, influencing recovery, loading and crushing costs and overall production. Different blast designs have been noticed to influence fragmentation, however, measuring it has been difficult and costly. The flow and draw body in sublevel caving are greatly influenced by the fragmentation. Due to the confined space, however, a key fragmentation process that occurs during blasting is missing. Therefore, knowledge on how to optimise fragmentation through different blast designs, leading to lower costs and improved productivity is paramount. Using a software based on 3D fragmentation analysis, a contactless data capturing was undertaken by two cameras mounted on the roof of the drift, collecting data without disturbing production. A total of 5029 bucket loads of load-haul-dump machines have been analysed from 15 rings and, based on the results, two initiation systems compared. The results were also collated to previous studies from the mine with a burden of 3 m, however, a direct comparison was hindered by a lack of data. The two analysed initiation systems are the test system consists of the nonelectric system and the reference system. In the reference system currently used by LKAB, the initiation system consists of a low grain detonating cord used to carry the initiation signal to the primers, that are coupled to nonelectric detonators with 60 cm long shock tubes. Whereas the test initiation system consists of a long shock tube directly coupled to the primer. The results show that the test system performs better than the reference system. On average, the number of boulders in the rings initiated according to the test system was lower by 81,6% compared to the reference rings. In the rings initiated with the test system, the x20 fragment size was improved (smaller fragment sizes) by 65,5%, in case of x50 by 32,1% and in case of x80 by 26,3%.