FluidMotion Robotics

Fail Operational AI-Powered Autonomy for Earth’s Most Critical Missions

Unified Autonomy platform for Air (UAV), Ground (UGV), and Water (USV) operations where failure is not an option

Waterways and Airspace Crises

Environmental

500+ tons of daily plastic | €200M+ economic impact

Safety

3,500+ annual US drownings | 70% lifeguard shortage

Defense & Surveillance

Rising waterway closures | Asymmetric maritime threats

Fluid Motion Robotics (FMR)

FMR is a California based company with close Greek business, engineering and experimental deployment ties. It is rooted in two strategically important coastlines, the Pacific Ocean and the Aegean and Mediterranean Sea. These regions demand resilient, autonomous systems for monitoring, response, and continuous operations across complex environments. FMR mission is to build the hardware and software to meet those demands.

Direction

Vision, mission, and execution roadmap

01

The Vision

Restore the world's most critical water and land infrastructures.

02

The Mission

  • Deploy water cleaning, airspace and land monitoring at scale.
  • Enable GPS-denied operations for UAVs and USVs with onboard AI decision making.
  • Surpass market leaders in performance.
03

The Goals 18 months

  • Commercialize DPilot Control and DPilot Horizon AI integration.
  • TFV manufacturability, robust DPilot Surf, and mini-scale demos.
  • Operations and strategic milestones across the USA and Greece.

Projects

One autonomy stack. Multiple mission profiles

FMR is developing deployable robotic systems and autonomy infrastructure for maritime operations, fail-safe vehicle control, and software-defined mission execution.

Water / USV

FMR Marine USV

Fully autonomous system End-to-end mission execution with zero human intervention during operation.
Computer vision Real-time detection and environmental awareness for floating debris and mission targets.
Fail-safe design Redundant systems engineered for reliable operation in critical marine environments.
Modular architecture Adaptable payload and mission design for cleanup, SAR, inspection, and defense use cases.
Embedded Controls

DPilot Controls; Baseboard

Fail-operational controls Designed for reliable embedded autonomy.
Cross-domain scalability Scalable platform for USV marine, UAV aerial, and UGV ground robotic systems.
Power Integrity

DPilot Controls; Power Module

No Single Point Failure (SPF) design Architected to avoid single points of failure across the power system.
Dual power inputs Redundant battery sources maintain system availability during abnormal power events.
Power redundancy Redundant high-current and low-current paths for propulsion and control electronics.
Onboard monitoring Real-time status, voltage, current, and temperature monitoring for system health visibility.
Software Platform

DPilot Horizon

Semi-autonomous operation Operators select the route while missions stream to the vehicle in flight, enabling live route edits.
Localization EKF fuses IMU, compass, and GPS into a continuous estimate of position, velocity, and attitude.
Dead reckoning IMU-based estimation maintains continuity between GPS updates.
Motion control Hybrid authority architecture blends operator commands with onboard stabilization and flight controllers.

DPilot in action

ROS2 closed-loop simulation of Trash Feaster with Nav2, RViz, and DPilot waypoint mission coverage

ROS2 Closed-Loop Simulation (Nav2 + RViz + Coverage)

This simulation runs the Trash Feaster mission in closed loop using ROS2, Nav2, RViz, and a coverage algorithm implementing DPilot waypoint missions. In closed loop, the controller continuously reads state feedback and updates commands during execution.

Isaac Sim open-loop physics simulation of Trash Feaster vehicle

Isaac Sim Open-Loop Physics Simulation

This NVidia Isaac Sim clip focuses on the Trash Feaster vehicle physics in open loop. In open loop, command inputs are applied without mission-level feedback correction, which is useful for isolating and validating baseline dynamics.

DPilot UAV real flight mission debrief — manual take-off, auto cruising, mixed mode landing at Sunnyvale Baylands

Real Flight Mission — Sunnyvale Baylands, April 2026

Annotated debrief of a DPilot UAV flight at Baylands Park, Sunnyvale, California. Three segments are shown: manual take-off, autonomous cruising mode, and mixed-mode landing. Log replay functionality that enables the development of localization, motion planning and motion control (flight control) algorithms using logged test data.

DPilot Surfer YOLO TensorRT trash detection test clip

DPilot Surfer Vision — Jetson TensorRT Test

Custom YOLO/SSD model trained on Taco datasets and accelerated by TensorRT on Jetson hardware. A Zed 2i stereo Camera is used for depth estimation. Limited lighting tests stressing trash detection.

Team

Built by engineers

A multidisciplinary team focused on hardware, systems, software, and controls — built for real needs and real applications.

Contact / Investment

Partner with FluidMotion Robotics

We are looking for pilot partners, strategic collaborators, and investors aligned with the future of fail-operational autonomous systems.

Submissions are securely handled through Formspree and sent to the FluidMotion Robotics team.