Dedrone’s 2025 intelligence report, published to coincide with DSEI, is a useful reality check for anyone building or buying counter-UAS systems. The report compiles operational detections across Dedrone’s Europe, Middle East, and Asia sensor network and highlights three trends that matter for fielded security programs: a fast rise in DIY builds, a meaningful share of low-visibility and night operations, and a decline in dominance by a single manufacturer. These shifts change how defenders should design detection mixes and response playbooks.

Headline numbers to lock in. Dedrone reports a roughly 4.3x increase in detected DIY drones in 2025 versus 2024 and that 37.5 percent of 2025 detections to date happened in low-visibility conditions. The company also shows DJI’s share of detections falling from about 95 percent in 2024 to roughly 83 percent in 2025. Treat these as vendor reported metrics, but treat their operational implications as real: more diverse airframes plus more night and RF-silent activity equals higher false negatives for RF-first defenses.

Tactics, not just technology. The white paper catalogs tactics that undermine single-modality systems: RF-silent flights, GPS-free navigation using SLAM and visual odometry, fiber-optic tethered platforms that defeat jammers, and composite attacks where swarms or decoys saturate sensors while higher value elements move in. Dedrone frames modern attacks as multi-domain and modular. If you are responsible for an airport, stadium, or critical asset, assume adversaries will mix low-cost commercial parts with bespoke navigation tricks long before you see them at scale.

Real world validation at DSEI. Dedrone’s deployed RapidResponse systems at DSEI 2025 recorded continuous activity across the event window. Their event reporting shows dozens of flights detected, multiple night-time alerts, and police responses triggered by Dedrone detections. That kind of exercise matters because it tests detection-to-decision timelines in complex crowded environments. For practitioners, the lesson is straightforward: run operational exercises with multi-sensor feeds and the people who will act on them, not just vendors who demo single sensors.

What to change in procurement and lab work. First, prioritize sensor fusion. RF remains high value for broad coverage but fuse it with wideband radar, EO/IR, acoustic, and advanced ML on the decision side to catch RF-silent or spoofed flights. Second, invest in software-first architectures that allow incremental sensor swaps. The report argues for software-defined defenses that can adapt by updating classifiers and ingesting new signatures rather than ripping hardware out. Third, plan your engagement rules around classification confidence and risk prioritization. When attackers use swarms or cheap decoys, you must reserve kinetic or high-cost mitigations for the highest risk tracks.

Quick, practical steps you can implement in a lab this quarter. 1) Build a three-sensor fusion testbed: a cheap RF monitor, a small FMCW radar or hobbyist radar board, and a PTZ EO camera. Stream those feeds into a local instance of an open inference engine and label simulated flights. 2) Emulate RF-silent flights by toggling RF emission on a test UAV and evaluate detection degradation. 3) Run night and clutter tests. Dedrone’s data shows many detections occur in low-visibility settings, so verify your imaging and ML stacks under those conditions. 4) Exercise escalation: simulate an incursion and run through alerting, operator classification, and the chain of response you will use with law enforcement or security teams. These are low-cost, high-value experiments.

The economics and policy angle. Dedrone quotes an uncomfortable math problem: low-cost attack drones can trigger very high-cost responses from legacy systems, making the asymmetry intolerable at scale. That is why cost-matched mitigation strategies and automation matter. Procurement teams need to balance price per defended kilometer, software upgrade paths, and operational costs for people. Also keep an eye on legal and regulatory constraints when you test mitigations; non-kinetic and kinetic actions have different approvals and liability profiles.

Bottom line. Dedrone’s 2025 intelligence output should not be treated as a vendor marketing brief, but it does offer credible operational signals: a more diverse drone ecosystem, more activity in low-visibility windows, and tactics that erode RF-only defenses. If you run a security lab or lead a procurement program, prioritize sensor fusion, software-first architectures, and realistic exercises that stress detection and decision chains. That combination buys you the adaptability the report says will decide future airspace security outcomes.