Taiwan's audit finds a $111M counter-drone system that cannot hear the band the PLA moved to

Taiwan's Audit Office has reported that a $111 million fixed-site counter-drone programme, developed by the National Chung-Shan Institute of Science and Technology to protect airbases and other military facilities, cannot detect or jam the 5.1GHz frequency used by newer PLA drone models ^[1][2]. The finding arrived shortly after the army terminated a separate purchase of 26 fixed-site counter-drone systems from local contractor Tron Future Tech, which failed its third and final acceptance test ^[3][4], which means both of Taiwan's fixed-site lines are in trouble at the same time ^[14].

The audit's list of gaps is not one missed frequency. The system also does not fully cover parts of the BeiDou-3 satellite navigation constellation, and has limited capability against fibre-optic and AI-guided drones ^[5]. On top of the capability findings, the programme was already six to twelve months behind schedule because of construction delays and shortages of key components, and the audit warned the combination could undermine protection for military bases ^[6][7].

The defence ministry's answer is a scheduling answer, and it is the most informative part of the story. The ministry said many of the technologies the audit flags emerged after the counter-drone programme began in 2021 ^[8]. Spokesman Sun Li-fang said NCSIST has already verified jamming for the newer frequency and will introduce it through future upgrades, and that the ministry is accelerating work on systems able to handle fibre-optic and AI-guided drones ^[9][10]. Read plainly: the requirement was frozen in 2021, the threat was not, and the fix has to travel back through the same programme that is already six to twelve months late ^[6][8][9].

The Tron Future Tech termination shows the other failure mode, where the gate works but the product does not clear it. Defence minister Wellington Koo Li-hsiung said the system struggled to combine information from search and passive radar when facing multiple waves of drones, making accurate target identification difficult for operators ^[4]. That is a sensor fusion problem under saturation, which is precisely the condition a fixed site would be attacked in, and it is the kind of problem that only shows up in a hard acceptance test rather than a datasheet.

Analysts quoted in the report push past the individual programmes. Shu Hsiao-huang of the Institute for National Defence and Security Research said Taiwan should stop treating every unmanned aircraft as the same threat, and warned that reliance on electronic jamming will lose value as fibre-optic and AI-guided drones cut their dependence on radio-frequency links ^[11][12]. Zivon Wang of the Chinese Council of Advanced Policy Studies argued Taiwan needs a cheaper kill, because firing expensive missiles at inexpensive drones becomes economically unsustainable ^[13]. Officials say upgrades are in development; analysts question whether they arrive fast enough ^[15].

Three things to watch. Whether the 5.1GHz jamming capability ships as a retrofit to already-built sites or only in a later block, since the difference decides whether the $111 million buys current coverage ^[1][9]. Whether the army recompetes the 26-unit requirement, and whether the new acceptance test explicitly scores multi-wave fusion rather than single-target detection ^[3][4]. And whether any of the promised layering includes a low-cost interceptor, because the cost exchange argument does not improve with better radar ^[13].