A modular engine for routing, simulation, and execution of distributed unmanned operations, FLOE transforms drone mission planning from a disconnected, multi–tool process into a unified operational workflow that helps users move from scenario definition to executable mission output with greater speed, transparency, and operational confidence.
Executive Summary
The Fleet Logistics Optimization Engine (FLOE), now nearing completion under Phase II SBIR development, is an end-to-end UAS routing and mission planning platform designed to move users from mission need to mission-ready output within a single integrated system. Rather than functioning as a standalone routing algorithm, FLOE provides a complete operational workflow that connects scenario development, routing, mission generation, simulation, visualization, and export.
FLOE supports many-to-many routing scenarios in which multiple launch or deployment locations can be matched against multiple targets, delivery points, areas of interest, or task locations, then organized into feasible routes based on realistic operational limits. The platform is built to accommodate multiple mission structures and routing conditions, including point-to-point visitation, delivery-style hub-and-spoke operations, polygon coverage missions, deployment location selection, return-to-home operational patterns, and roadway-constrained routing for dense urban environments.
A central strength of FLOE is that it is designed to generate routes that are operationally usable, not merely mathematically optimal. The platform accounts for practical considerations such as aircraft endurance, altitude transitions, loiter time, reserve margins, vehicle-specific performance parameters, and mission structure requirements. It gives users a straightforward way to evaluate operational options, assign work across available assets, and generate mission outputs that can be used directly in downstream simulation and execution environments.
FLOE is also designed as an integrated, user-centric system. Users can define scenarios manually or upload them from external sources, incorporate awareness of restricted airspace and obstacles, select or manage UAS profiles, generate routes, compile routes into executable mission plans, simulate those missions, visualize them in an interactive 3D environment, and export mission-ready files for autopilot systems or analysis workflows. This unified workflow reduces the fragmentation often associated with drone mission planning, where users otherwise must rely on multiple disconnected tools for design, routing, simulation, and export.
The platform already reflects use cases across diverse operational environments, including urban delivery, agricultural survey, coastal search and rescue, and disaster response. It also supports multi-flight dispatch when a single route is insufficient to service all tasks, making it suitable for scalable and repeatable UAS operations.
At its current maturity level, FLOE is best understood as a flexible, easy-to-use UAS operations platform that brings planning, routing, mission generation, simulation, visualization, and export together into a single end-to-end capability.
About Aerial Vantage
Aerial Vantage is a specialized drone technology and geospatial intelligence company founded in 2021. It combines decades of expertise in manned and unmanned aviation with advanced AI, machine learning, and computer vision capabilities to deliver end-to-end aerial data solutions. The company operates as a “drone airline,” providing scalable data acquisition and analytics that turn raw drone imagery into actionable business insights.
Originally based in the Washington, D.C. area, with ties to Leesburg, Virginia through parent and affiliate relationships with Mosaic ATM, the company relocated its corporate headquarters to the MSU Technology Innovation Center in East Lansing, Michigan in July 2023. As a private company with a small, agile team, Aerial Vantage focuses on automating drone operations and maximizing return on investment through efficient, data-driven decision support.
Aerial Vantage’s flagship offerings center on its Accelerate SaaS geospatial intelligence platform, which processes, analyzes, stores, and presents drone-acquired data using AI-driven tools. Key platform features include polygon-based areas of interest, project and run workflows for repeatable multi-temporal analysis, metadata tracking for provenance, secure cloud or on-premise storage, automated data extraction from multiple sensor types, a user-friendly web interface, flexible reporting, and real-time geospatial alerts.
Complementary services include Drone Flights as a Service, including FAA-approved Beyond Visual Line of Sight (BVLOS) waivers for large-scale operations, the Fleet Logistics Optimization Engine (FLOE) for automated mission planning and routing, and custom advanced analytics such as crop yield forecasting, anomaly detection, spectral imaging, and geospatial classification.
These capabilities allow Aerial Vantage to deliver both the “flights” and the “insights” through a unified operational model. The company serves agriculture, energy, natural resources, and municipal or government customers, with a focus on helping organizations scale drone operations safely, efficiently, and affordably.
Within this portfolio, FLOE represents Aerial Vantage’s mission-planning and operations layer: the system that converts mission needs into structured, feasible, and mission-ready flight outputs.
FLOE: Platform Overview
The Fleet Logistics Optimization Engine (FLOE) is a modular platform for planning, optimizing, validating, and exporting unmanned system operations across a range of mission types. At its core, FLOE provides a unified environment for transforming geospatial inputs such as launch facilities, targets, delivery points, survey polygons, and operational constraints into executable mission plans.
FLOE is designed to support many-to-many routing across multiple operational contexts. Multiple launch or recovery points can be evaluated against multiple demand locations, areas requiring coverage, or mission tasks, and the resulting assignments can be organized into feasible routes according to vehicle range, battery limits, altitude transitions, loiter requirements, and mission-specific logic.
The platform is built around several principal operational use cases:
- Deployment location selection
- Point-to-point routing
- Coverage and survey routing
- Delivery-style hub-and-spoke routing
- Return-to-home mission structures
- Roadway-constrained urban routing
Rather than treating routing as an isolated optimization problem, FLOE integrates route creation into a larger operational workflow. Users can define or import a scenario, assign vehicle parameters, evaluate constraints, generate route candidates, compile those routes into mission plans, simulate them, visualize them in a 3D environment, and export outputs into mission-ready formats.
Core Capabilities
Scenario Development
FLOE allows users to create or upload scenarios that include launch points, target locations, survey polygons, and environmental or regulatory context. Operators can use freeform map-based editing tools or structured geospatial inputs to build mission scenarios quickly without relying on external GIS software.
Deployment Determination
Routing Algorithms
FLOE includes separate tabs for wrapping UAS routers for different use cases. This includes point-to-point visitation, coverage/survey routing, return-to-home delivery style routing, and special roadway routing for operation in highly congested urban areas along roadway corridors. Accounts for UAS parameters, loiter time, altitude, and other considerations.
FLOE includes multiple routing modules tailored to distinct mission types:
- point-to-point visitation routing,
- coverage routing over polygons,
- hub-and-spoke delivery routing,
- return-to-home facility missions, and
- roadway-constrained routing for urban operations.
These routing capabilities account for aircraft-specific operational considerations such as loiter time, battery reserves, cruise behavior, climb and descent phases, and altitude conventions.
End-to-End Simulation and Export
FLOE extends beyond route generation into mission compilation and simulation. Routes can be converted into executable flight plans, simulated in either fast kinematic or higher-fidelity environments, visualized in 3D, and exported into autopilot-ready formats for downstream execution.
User-Centric Design
The platform is intentionally designed to be approachable for operational users while remaining technically robust. It includes comprehensive documentation, an expanding library of prebuilt vehicle profiles, and the ability for users to define new platforms. Simplified yet feature-rich environment with comprehensive user-facing documentation. Currently building up a pre-existing library of common UAS platforms and ability to create more.
Capability Status and Functional Detail
FLOE’s maturity reflects a system that has moved well beyond conceptual design and into operational capability. The table below is expressed here in narrative form to support white paper readability while still documenting current status.
Deployment Location Selection, Operational
FLOE includes an operational greedy max–coverage set cover solver that determines the minimum set of launch and recovery facilities needed to place all target locations within round–trip UAS range. The capability includes optional OpenStreetMap road–snapping to improve crew accessibility and a configurable cap on the number of facilities allowed in the solution. The underlying method is an FLP–based greedy set cover approach.
Point-to-Point Routing, Operational
The point–to–point routing module is a multi–stop, multi–sortie optimizer that builds routes from deployment facilities to distributed target locations. It supports multiple flights per facility in order to maximize coverage within endurance limits and uses Voronoi–based demand assignment alongside battery and reserve modeling. The underlying routing logic is based on nearest–neighbor TSP methods coupled with Voronoi assignment.
Coverage Routing, Operational
FLOE supports systematic survey route generation over user–defined area polygons. This capability includes configurable camera field–of–view and overlap settings, automatic band–splitting for endurance–aware multi–flight execution, and two distinct routing strategies: standard boustrophedon lawnmower coverage and an adaptive lawnmower approach that better handles concave geometries. The underlying logic is strip–based coverage generation with endurance band splitting.
Delivery Routing, Operational
The delivery routing module supports hub–and–spoke, round–trip mission structures appropriate for per–delivery missions. It models the full altitude transition cycle, including climb from the hub, descent at the delivery site, climb at the delivery site, and descent on return, along with ground delivery time. This results in more realistic range and battery calculations than simplified straight–line delivery models. The capability uses Voronoi hub assignment combined with per–demand range gating.
Return-to-Home / Facility Missions, Operational
FLOE supports facility–to–task–to–facility routing for operational patterns in which the aircraft must return to its origin after completing a task sequence. This is important for recurring operational workflows where launch and recovery must occur at the same site for each sortie.
Urban Roadway-Constrained Routing, In Development
A roadway–constrained routing capability is currently in development to support operations in dense urban environments or scenarios that require alignment with road corridors or ground–vehicle–adjacent operations. This capability is being designed to incorporate OSM road graph data to enforce street–network adherence.
Live FAA Airspace Integration, Operational
FLOE includes live advisory overlays for three FAA data sources:
- controlled airspace, including Class B, C, D, and E2 areas that may require LAANC or DroneZone authorization,
- National Security UAS Flight Restrictions (NSUFR) that function as hard no-fly zones, and
- FAA Digital Obstacle File (DOF) obstacles filtered by above-ground-level height.
These datasets are queried live against the current map viewport, and generated routes can be flagged on export when airspace conflicts are detected.
Mission Compile-Simulate-Export Pipeline, Operational
FLOE includes a three-step pipeline that converts routed GeoJSON into mission-ready files.
Compile cleans route geometry, injects takeoff, climb, and landing waypoints, and assigns altitude conventions and speed profiles.
Simulate supports either fast kinematic interpolation or higher-fidelity PX4 + JSBSim software-in-the-loop simulation.
Export produces mission artifacts such as PX4/QGroundControl mission.json, QGC WPL 110 waypoint files, and trajectory_reference.csv for analysis.
PX4 + JSBSim SITL Integration, Operational
For higher–fidelity validation, FLOE supports PX4 autopilot software–in–the–loop with JSBSim 6DoF flight physics. The system supports both quadrotor and fixed–wing airframes, manages process lifecycle, parses ULG flight logs, and returns trajectory outputs for analysis. It also supports QGroundControl debugging via localhost UDP.
Auto-Pilot Ready Exports, In Development
FLOE is expanding export support for enterprise autopilot ecosystems, including formats such as DJI Enterprise-compatible outputs. The goal is to allow users to export final routes and move directly into execution with minimal additional formatting.
3D Mission Visualization (CesiumJS), Operational
The Play Scenario environment provides interactive mission playback on a 3D globe. It supports drone models, glowing trail visualization, OSM–based 3D buildings for urban context, a scrubbable timeline, and playback speed control from 1x to 60x. It accepts GeoJSON, mission files, waypoint files, and trajectory CSV inputs.
Scenario Builder, Operational
FLOE includes a freeform scenario builder for constructing mission inputs without external GIS tooling. Users can place facilities, demand points, and polygons directly on the map, edit vertices interactively, paste bulk coordinates, and export scenarios directly into routing workflows.
Modular Architecture, Operational
Each routing capability is implemented as an independently deployable module, typically consisting of a Python FastAPI backend paired with a React frontend tab. Each module maintains its own schema, optimizer, standalone runner, and test suite. This architecture allows algorithms to be swapped without structural disruption and supports full Docker orchestration with configurable ports and data paths.
System Architecture
FLOE is built on a modular, service–oriented architecture that separates user interaction, planning logic, simulation, and export functions into distinct but interoperable layers. This design supports extensibility, repeatability, and integration with both user–facing workflows and external operational systems.
Application Layer
The front end is a web-based interface organized around mission-planning workflows and discrete capability tabs. Users interact with FLOE through map-based editing, parameter selection, routing controls, and scenario visualization tools. Key frontend responsibilities include:
- scenario creation and editing,
- configuration of vehicle and mission parameters,
- visualization of route candidates and constraints,
- execution control, and
- display of metrics and diagnostics.
Service Layer
The backend consists primarily of Python services, including FastAPI-based modules that expose each routing or mission-processing capability through defined interfaces. This supports both interactive use through the UI and programmatic integration into external workflows. The backend handles:
- routing and optimization algorithms,
- feasibility and constraint checks,
- mission compilation,
- simulation orchestration,
- data transformation, and
- standardized output generation.
Data and Integration Layer
FLOE operates on structured geospatial inputs and standardized mission artifacts. Common data structures include facilities, demand points, areas of interest, route geometries, environmental overlays, and export files. By relying on common geospatial and mission formats such as GeoJSON and autopilot–specific mission files, FLOE can interface with GIS tools, simulation environments, and ground control systems.
Execution Modes
The platform supports several operating modes:
- Interactive mode for end users performing scenario planning and mission design through the web interface,
- API mode for integration into automated pipelines or other applications, and
- Standalone mode for benchmarking, algorithm testing, or offline execution.
This flexibility allows FLOE to function both as a user-facing mission-planning environment and as a backend engine within broader autonomy or operations workflows.
Design Principles
FLOE’s architecture is guided by the following principles:
- Modularity: Capabilities remain independently deployable and extensible,
- Transparency: Diagnostics and performance indicators are visible to users,
- Interoperability: Data formats and interfaces support downstream integration,
- Scalability: Workflows support both small scenarios and large, distributed operations,
- Operator-Centricity: The system is organized around real-world planning tasks rather than abstract optimization alone.
Operational Workflows
FLOE is designed around a mission–centric workflow that guides users from operational concept through mission–ready output. The workflow emphasizes rapid iteration, visibility into feasibility and tradeoffs, and continuity between planning stages.
1. Scenario Definition
Users begin by defining the operational environment. This may include launch and recovery facilities, delivery or visitation points, survey polygons, and context such as airspace, obstacles, or road accessibility. Scenarios can be created directly in the platform or imported from external datasets.
2. Routing and Optimization
Users select the routing capability that best fits the mission type. FLOE then applies the appropriate optimization logic while accounting for aircraft endurance, reserve requirements, mission structure, and operational constraints. At this stage, the system:
- assigns demand to available facilities,
- determines reachability,
- generates feasible route geometries,
- surfaces infeasible tasks or orphaned demand, and
- provides route-level metrics for comparison.
3. Mission Generation
Optimized routes are then compiled into executable mission structures. This includes waypoint injection, altitude and speed profile assignment, and normalization of geometry into a mission representation that can be simulated or exported.
4. Simulation and Validation
Users can validate mission timing, sequencing, and behavior using either lightweight kinematic simulation or higher–fidelity PX4 + JSBSim simulation. The 3D visualization environment gives operators an intuitive view of route geometry, timing, and environmental context.
5. Export and Integration
Once validated, missions can be exported into formats appropriate for autopilot systems, analysis pipelines, or downstream operational tools. This allows FLOE to bridge the gap between planning and execution without requiring redundant manual conversion steps.
6. Iterative Planning
A major advantage of FLOE is that the workflow is iterative. Users can adjust facilities, route parameters, vehicle assumptions, or areas of interest and rerun optimization quickly. This supports trade–space exploration and makes it easier to converge on operationally viable mission plans.
The Florida NextGen Test Bed
Established in 2008, the Florida NextGen Test Bed (FTB) is an FAA-sponsored research and integration environment operated by Embry-Riddle Aeronautical University under an Other Transaction Agreement. The FTB is designed to support the evaluation of emerging air traffic management concepts, technologies, and operational procedures within a high-fidelity, representative National Airspace System environment without affecting live operations.
The FTB enables collaboration among government, industry, and academia. Its infrastructure spans en route automation, terminal operations, surface management, oceanic operations, and traffic flow management. This breadth makes it possible to evaluate both individual technologies and their interaction within larger system-of-systems contexts.
The environment combines simulation and real-world system integration, incorporating representative or fielded FAA platforms such as ERAM, STARS, and TBFM, along with research tools and decision-support technologies. Dedicated hardware, cybersecurity enclaves, and private network co-location capabilities make it possible to integrate external systems into realistic end-to-end test environments.
The FTB also supports human-in-the-loop experimentation and real-time simulation, allowing researchers to examine not only system performance but also operator interaction, workload, and decision-making.
As evaluations expand to include advanced air mobility, distributed autonomy, and UAS operations, the complexity of test scenarios continues to increase. These scenarios often involve coordination across multiple vehicles, facilities, and constraints while accounting for endurance, sequencing, airspace structure, and mission dependencies.
In that context, there is a growing need for tools that can rapidly build, adapt, and evaluate mission scenarios before they are introduced into a high-fidelity NAS environment. FLOE is well positioned to support this need by translating high-level operational ideas into structured, repeatable, and executable mission constructs that can then be evaluated within environments like the FTB.
Strategic Fit and Value Proposition
The Florida NextGen Test Bed is designed to evaluate emerging operational concepts within a realistic representation of the NAS. As research expands to include autonomous systems, UAS logistics, and AAM operations, flexible scenario planning and mission construction tools become increasingly important. FLOE aligns directly with this requirement by providing a modular planning and optimization layer that complements, rather than replaces, existing simulation and integration infrastructure.
Strategic Fit
Rapid Concept Evaluation
FLOE helps researchers and operators move quickly from mission idea to structured, testable scenario.
System-of-Systems Integration
The platform provides a consistent way to represent multi-vehicle, multi-facility, and multi-mission operations that can feed larger simulation environments.
Repeatable Experimentation
Standardized inputs and outputs make it easier to recreate, compare, and refine mission scenarios across repeated test runs.
Scalability Across Mission Types
FLOE supports multiple operational paradigms, from centralized delivery routing to distributed survey and visitation tasks, without requiring platform redesign.
Support for Emerging Operations
The system is well suited to exploratory work in AAM, UTM, UAS logistics, hybrid operations, and mission-level autonomy.
Value Proposition
Bridging Concept and Execution
FLOE creates a structured path from abstract operational intent to executable mission artifacts.
Improved Trade-Space Exploration
By exposing metrics related to coverage, time, resource utilization, and infeasibility, FLOE helps users compare operational strategies more rigorously.
Reduced Integration Overhead
Standard geospatial outputs and autopilot-ready mission artifacts reduce the friction associated with moving plans into simulation or execution environments.
Enhanced Transparency and Diagnostics
The platform surfaces route assumptions, feasibility limits, and performance metrics so users can understand both results and constraints.
Accelerated Iteration Cycles
Because scenarios can be built, rerouted, simulated, and exported within a single environment, users can evaluate more concepts in less time.
Together, these strengths position FLOE as a practical enabling technology for research, experimentation, and operational planning in increasingly complex unmanned and autonomous mission environments.
Conclusion
The Fleet Logistics Optimization Engine represents a significant step forward in how UAS missions can be planned, evaluated, and operationalized. As drone operations become more distributed, complex, and mission-critical, organizations need more than isolated routing algorithms or disconnected planning tools. They need integrated systems that can translate mission intent into feasible, mission-ready outputs while preserving visibility into constraints, tradeoffs, and execution requirements.
FLOE addresses that need by bringing scenario development, deployment selection, routing, mission compilation, simulation, visualization, and export into a single unified platform. Its architecture is modular enough to support multiple mission types and evolving algorithms, while its workflow is practical enough to serve real operational users. This combination of technical flexibility and operational usability is one of FLOE’s strongest differentiators.
The platform’s current capabilities already demonstrate applicability across a broad range of use cases, including delivery operations, area survey, search and rescue, disaster response, and urban mission planning. Its support for multi-sortie operations, airspace awareness, autopilot-ready exports, and high-fidelity simulation makes it especially valuable for organizations seeking not only to optimize missions on paper, but to prepare them for realistic execution and evaluation.
For environments such as the Florida NextGen Test Bed, FLOE offers a valuable bridge between concept development and system-level assessment. It enables mission scenarios to be built quickly, tested consistently, and refined iteratively, helping researchers and operators explore emerging UAS and AAM concepts with greater rigor and efficiency.
As development nears completion under Phase II SBIR, FLOE is positioned not just as a routing tool, but as a scalable UAS operations platform—one that supports the full lifecycle of mission planning from initial scenario design to mission-ready output. In that sense, FLOE is not simply about finding routes; it is about enabling operationally grounded autonomy.