We scan new podcasts and send you the top 5 insights daily.
Tenet Industries' CEO asserts that "last mile" targeting for drones is a solved problem from the 1960s. The real innovation and demand lie in enabling drones to navigate autonomously for 20+ kilometers in GPS-jammed environments, a much harder and more critical problem to solve.
Even the simplest form of drone AI—terminal guidance, where the AI takes over for the final 500 meters—had a massive impact. One pilot's precision mission success rate more than tripled, and his effective 'kill zone' expanded from 3km to 10km, demonstrating AI's immediate battlefield value.
The universal jamming of GPS in conflict zones like Ukraine has rendered modern, GPS-reliant drones ineffective. This has reset the baseline level of drone autonomy to what it was two decades ago, creating an urgent need for alternative navigation systems that can operate without satellite signals.
Skydio's CEO details the drone industry's progression: from raw stick-to-motor commands, to microprocessor-stabilized attitude control, to GPS position hold, and finally to AI-driven computer vision. Each step abstracted low-level flight control, making drones progressively more accessible and autonomous.
Theseus's vision-based navigation is only accurate to 30 meters, a deliberate choice. This is sufficient for long-range transport ('getting from A to B') without enabling precision targeting. This strategy prioritizes reliability in GPS-denied areas while navigating regulatory and ethical concerns.
Skydio's drones are designed as 'force multipliers' where AI handles complex tasks like navigation, obstacle avoidance, and subject tracking. This frees the human operator to focus on high-level mission objectives, like assessing a situation, rather than the mechanics of flying the drone.
The intense signal jamming by Russia in Ukraine makes remotely piloted drones ineffective in the final phase of an attack. This has created a tactical necessity for drones that can autonomously complete their mission after losing their data link, accelerating the development of practical, on-board AI for target engagement.
Inspired by self-driving cars, a framework for drone autonomy has emerged: L1 (Terminal Guidance), L2 (Bombing), L3 (Target Detection/Engagement), L4 (Navigation), and L5 (Takeoff/Landing). This provides a clear roadmap for developing and classifying autonomous capabilities on the battlefield.
An FPV drone is already three orders of magnitude more versatile than an artillery shell. Adding full autonomy adds another *four* orders of magnitude in capability by expanding the user base (100x), increasing mission success (10x), and improving utility per drone (10x).
Ukraine is pioneering 'last mile autonomy' not as a strategic push for automation, but as a tactical necessity. When Russia jams the data link to a drone, the system can autonomously complete the final leg of its attack on a pre-identified target, countering electronic warfare.
The war in Ukraine has evolved from a traditional territorial conflict into a "robot war," with drones dominating the front lines. This real-world battlefield is accelerating innovation at an "unbelievable" pace, driving new solutions for secure communications and autonomous targeting, providing critical lessons for US drone strategy.