Ai tracking integration drones and Betaflight Jetson pi hd zero analog nav

Position: ai tracking integration with drones and Beta flight Jetson pi hd zero analog nav
Location: California

AI Tracking + FPV Overlay Integration Engineer (Beta flight + Jetson/Raspberry Pi + HDZero/Analog)

We’re hiring a contractor to integrate real‑time AI target detection/tracking into FPV drone systems, with a low‑latency on‑screen overlay (lock box/reticle/cues) and optional track‑assisted flight behaviors. You’ll work across Beta flight, MSP Display Port/OSD, HDZero (digital), and analog video paths, using a companion computer (NVIDIA Jetson and/or Raspberry Pi).

What you’ll build

• Real‑time object detection + tracking on a live FPV feed
• Dynamic bounding box / lock box overlay visible in goggles (HDZero OSD path and/or analog OSD path)
• Stable performance with graceful failure behavior (flight remains safe and fully pilotable if AI pauses / crashes)

• Track‑assisted control (e.g., yaw/pan cues, target centering assist, or autopilot‑like "hold target" mode)
• Click‑on‑center (toggle switch to lock target)
• Multiple targets identified
• "NAV" overlays (home arrow / heading / GPS data / target bearing cues), integrated cleanly with existing Beta flight OSD elements
• First deliverable does not need to control the drone – simply lock on targets and assist with identification of objects and targets.
• Recording pipeline with overlay burned‑in (DVR + post/real‑time encoded output)

Responsibilities

• Design the end‑to‑end architecture: camera/video ingest → AI inference → tracking → overlay rendering → FPV display
• Implement a robust overlay path using Beta flight MSP (e.g., MSP Display Port/Canvas) or an equivalent method that keeps latency minimal
• Integrate companion computer comms with the flight controller (UART/USB), including watchdog/health checks
• Optimize for real‑time performance (frame rate, inference speed, end‑to‑end latency)
• Produce reproducible setup docs and scripts (build steps, configs, wiring diagrams, parameters)
• Provide a demo showing: target acquisition, persistent track, overlay behavior, and fail‑safe behavior

Required experience / skills

• Strong Linux + embedded development experience
• Solid Python and/or C/C++ (OpenCV, tracking filters, threading/async, profiling)
• Real‑time video handling: GStreamer/FFmpeg, V4L2, CSI/USB cameras, hardware encode/decode
• Beta flight integration experience, specifically:
• MSP messaging and/or MSP Display Port / OSD overlay concepts
• Serial comms and telemetry handling
• Comfort working with Jetson (Orin family preferred) and/or Raspberry Pi (Pi 4/5)
• Ability to test on real hardware and iterate quickly from flight logs + DVR

Nice‑to‑have

• HDZero ecosystem familiarity (OSD expectations, latency constraints, typical wiring)
• Analog FPV overlay methods (OSD chips, FC OSD, external overlay options)
• Experience with UAV "navigation" overlays and sensor fusion (GPS/compass/IMU cues)
• Control systems experience (PID‑style target centering, rate limiting, smoothing)
• Familiarity with Ardu Pilot/iNav (only if we extend beyond Beta flight later)

Target hardware stack (typical)

• Beta flight flight controller (OSD‑capable)
• Companion computer:
  Jetson Orin Nano / Orin NX and/or Raspberry Pi 5
• Video: HDZero (digital) and/or analog FPV
• Camera(s): CSI/USB/HDMI capture depending on stack
• UART link between companion computer and FC (plus power/ground best practices)

Deliverables & acceptance criteria

• Working prototype on a real quad: overlay visible in goggles while flying
• Track box stays stable on a moving subject (with reasonable occlusion handling)
• Measured end‑to‑end latency and FPS (with your test methodology documented)
• Fail‑safe behavior: if AI stops, OSD/telemetry remains normal and pilot retains full control
• Clean repo + setup documentation + wiring diagram + config files
• Short demo video + brief technical walkthrough

How to apply

(send these)

• 2–4 examples of relevant work (video overlays, CV on edge devices, Beta flight/MSP work)
• Your recommended architecture for HDZero and analog (high level)
• What hardware you want us to provide (or what you already have)
• Your expected rate and availability

Seniority level

Entry level

Employment type

Contract

Job function

Project Management and Information Technology

Industries

Entertainment

#J-18808-Ljbffr