Weekly Digest — October 14, 2026

Counter-UAS activity intensified this week across both operational and research fronts. From record border incursion data to renewed base-defence contracts and advances in autonomous path planning, the field is moving on multiple tracks simultaneously.

Lead story

US Military Counts 378 Border Drone Incursions In 30 Days, And Rising — Joint Task Force Southern Border logged 378 drone incursions along the US-Mexico border within a single 30-day period, with detection rates still climbing. The Pentagon's response — Falcon Peak 26.2 at Yuma Proving Ground — is described as its largest and most complex counter-UAS exercise to date, stress-testing detection and defeat systems from 21 vendors against scenarios modelled on actual border threat patterns. The scale of both the incursion data and the exercise underscores how rapidly C-UAS evaluation requirements are outpacing existing frameworks.

In focus

• Nellis Air Force Base Renews UNIFY.C2 Contract Following Successful Year-Long Counter-UAS Evaluation — Nellis AFB has extended its contract with Spatial Persistent Software's UNIFY.C2 platform following a year-long operational evaluation, expanding coverage to Creech AFB and the Nevada Test and Training Range. The renewal adds a data point for vendor-neutral sensor fusion architectures in permanent base-defence deployments.
• 2DGS-Planner: How Rasterization Is Rewriting the Rules of Path Planning on Gaussian Maps — KAIST's Urban Robotics Lab has demonstrated a path-planning approach — 2DGS-Planner — that queries geometry directly from 2D Gaussian splatting maps through rasterization, bypassing the need to treat individual Gaussian primitives as discrete obstacles. Benchmark results show gains in roadmap connectivity, clearance accuracy, and overall planning success rates, with potential implications for UAS operating in map-constrained environments.
• When Every Detector Fails: MBZUAI's Calibrated Resynthesis Paradigm for Certifying Real Images — MBZUAI researchers found that all twenty deepfake detectors tested dropped below 2% accuracy under adversarial perturbation, and introduced a resynthesis-based certification method that bounds false authentications to no more than 1%. As drone-captured imagery increasingly feeds intelligence workflows, the rapid erosion of post-hoc verifiability carries direct relevance for UAS sensor data integrity.