SkyFlock UAViation Pvt. Ltd.

The Swarm
Is Intelligent Now

Foundational autonomous drone infrastructure — from simulation to real-world deployment. Planned, flown and analysed from one browser cockpit: flyMe.

STATUS: FIELD TESTING
FUNDING: MeitY · SISFS
INCUBATOR: IIT ROPAR
STACK: PX4 · ROS 2 · GAZEBO
// The Product

flyMe.
One Screen From Briefing to Touchdown.

Our browser-based ground control station: plan a survey on the map, fly it on PX4 hardware, watch live telemetry, and read the analysis — without leaving the cockpit.

Interactive walkthrough · demo telemetry ← Swipe the cockpit to explore →
flyme.services / fly
Ground Station
Drone 1 ⚡▾
Live Map Plan Manual
3D SimCameraCourse
N HDG 042°
ALT
30.2 m
T+ 00:04:12
POSITION · ARMED
Click the map to drop corners · 4 corners → survey plan · Drone 1 targeted
50 m
Link100%
ArmedYES
ModePosition
Battery87%
GPS14 sats · 3D
Altitude30.2 m
Speed4.1 m/s
HealthOK
Quick ActionsVehicleIntelligence
Arming
Flight
Preflight GO
Data linkGO
GPS / positionGO
BatteryGO
ArmableGO
Mission Plan Drone 1
Survey altitude30 m
Corridor width8 m
Overlap70 %
Camera triggerevery 4 m
Phases derived from 4 corners
  1. 01Takeoffto 30 m
  2. 02Transithome → corner A
  3. 03Survey6 legs · 412 m
  4. 04ReturnRTL · land at home
PX4 AUTO mission · 30 camera triggers · geofence 250 m
Mission Reports
Mission analysisFlight health (PX4)Building compliance
Flights 4
survey_ward7_0914Real analyzed
compliance_block_cReal analyzed
slalom_wide_06Sim analyzed
hover_quality_03Real analyzing…
Area covered1.84 ha
Mean path error0.42 m
Distance flown1,236 m
Flight time06:48
Drones1
Cross-track error along survey metres · 6 legs
1.51.00.50 mean 0.42 m
Mission phases 06:48 total
TakeoffTransitSurvey · 04:37RTL
Building compliance nadir survey · height MAE 0.33 m
BuildingMeasured heightBylaw limitStatus
Block C · plot 1411.6 m12.0 m✓ Compliant
Block C · plot 1513.9 m12.0 m▲ Exceeds by 1.9 m
Block C · plot 169.2 m12.0 m✓ Compliant
01 · Fly

Live map, real telemetry

Link health, arming, mode, battery, GPS, altitude and speed in one status bar. ARM, TAKEOFF, HOLD, RTL and LAND are one click — with a preflight GO/NO-GO gate.

02 · Plan

Drop corners, get a survey

Four corners on the map become a takeoff–transit–survey–return plan with camera triggers, uploaded to PX4 as a native AUTO mission.

03 · Safety

Failsafes that don't need the operator

Link-loss watchdog, geofence, battery-triggered RTL and an audit trail for every command — running independently of the UI.

04 · Reports

From flight log to evidence

Every flight becomes a report: coverage, path error, PX4 flight health, and building-height compliance with the measurement attached.

// Backed & Recognised

Funded by India's
Deep-Tech Ecosystem.

Government grant, seed capital and IIT incubation — SkyFlock is built on institutional backing, not just conviction.

Government Grant
₹6,00,000

MeitY Funding

A ₹6 lakh grant from the Ministry of Electronics & Information Technology, Government of India — backing SkyFlock's autonomous drone infrastructure.

Ministry of Electronics & IT · Government of India
Seed Funded
SISFS

Startup India Seed Fund

Raised seed funding and recognised under the Startup India Seed Fund Scheme — the Government of India's flagship programme for early-stage deep-tech ventures.

Startup India Seed Fund Scheme · DPIIT
Incubated
IIT Ropar

Incubated at IIT Ropar

Building from the incubation ecosystem of the Indian Institute of Technology Ropar — with access to faculty research, labs and deep-tech mentorship.

Indian Institute of Technology Ropar · Punjab
GrantMeitY · ₹6 Lakh
SeedStartup India Seed Fund
IncubatorIIT Ropar
FocusAutonomous UAV Infrastructure
// What We Build

One Platform.
From Simulation to Deployment.

flyMe Control Plane

The React + FastAPI ground control station operators use for mission planning, live telemetry, and deployment control. One pane of glass from briefing to touchdown.

React · FastAPI · WebSocket

v2_swarm Autonomy Runtime

The ROS2 Humble autonomy layer — MAPF/CBS path planning, PPO-trained collision avoidance, and a MAVSDK/PX4 hardware bridge. The intelligence that makes a fleet a swarm.

ROS2 Humble · Python · MAVSDK

Continuous Learning Loop

Every flight generates telemetry that feeds back into the training pipeline — automatic model evaluation, promotion, and deployment. The swarm gets smarter while you sleep.

PyTorch · MLflow · swarm_analytics
// Field Tests

Flown. Not Just Simulated.

Footage from SkyFlock field tests — real airframes, real weather, real telemetry.

Descent and landingCustom hexacopter · dusk field test
Takeoff, climb and hoverQuadcopter · open-field test
Return and touchdownHexacopter · golden-hour recovery
Climb-outHexacopter · sunset departure
// Why SkyFlock Is Different

Built as a Substrate,
Not a Demo.

01

Plugin-Extensible Architecture

Third-party mission behaviors and policies plug in without forking core code. Extend the platform, don't rebuild it.

02

Continuous ML Learning Loop

PPO policies auto-retrain from flight data and auto-promote when evaluation gates pass. No manual model shuffling.

v15live evasion policy
03

Honest Analytics

Ground-truth coverage computed from position logs — never timer-based proxies. Numbers you can stake a mission on.

04

Production Safety Stack

Geofencing, heartbeat monitoring, Return-To-Launch, and full audit trails — present from day one, not bolted on later.

05

Sim-to-Real Parity

The same stack that runs in Gazebo deploys to real Pixhawk and Pi 5 hardware. What you validate is what you fly.

96.5%success rate in simulation
// The Continuous Learning Lifecycle

Six Stages. One Feedback Loop.

Every flight makes the swarm smarter. The Base Feedback Layer connects all six stages into a single, continuously improving system.

◆ Base Feedback Layer
Real-time data from every stage feeds continuous model improvement — session identity, persistence, replay, and post-flight analytics span the entire lifecycle.
// Technology

The Stack That Flies.

Browser
React 18 · TypeScript · MapLibre
flyMe Backend
FastAPI · Python 3.10
v2_swarm
ROS2 Humble · MAVSDK
PX4 Firmware
Pixhawk Flight Hardware · Pi 5 Companion
◆ ML Arsenal
PPO Evasion Policy v1596.5% · 3-drone · 200 ep
YOLOv11n Object Detectionreal-time threat ID
TFT-BiLSTM Battery Predictionpre-emptive RTL
GNN Swarm Coordinatormulti-agent allocation

Built for extensibility, not for demos.

The SkyFlock substrate — mission registry, session identity, plugin SDK, and safety chain — took months to build and is the defensible core of the platform.

Every layer is independently testable, every mission is replayable, and every flight feeds the next training cycle. This is infrastructure, engineered to outlast any single mission type.

ROS2 FastAPI React PyTorch MAVSDK PX4 Gazebo stable-baselines3 MLflow
// About

Built in India.
Built for the World.

“
We are no longer proving autonomous swarm drones can work. We are proving this can become foundational autonomous drone infrastructure.
— SkyFlock UAViation, Haryana, India

SkyFlock UAViation Pvt. Ltd. is a deep-tech startup building foundational infrastructure for autonomous drone operations. We are based in India and focused on the hardest part of drone autonomy: not flying a single drone, but coordinating intelligent swarms that learn from every mission.

The SkyFlock platform is simulation-complete and flying on real hardware. Our autonomy stack runs 11 validated mission types in Gazebo, flyMe has commanded real PX4 aircraft in field tests, and our first customer pilot — drone-based building-compliance surveys for municipalities — is in preparation.

We are funded by a ₹6 lakh MeitY grant, seed-funded under the Startup India Seed Fund Scheme, and incubated at IIT Ropar.

We are looking for research collaborators, enterprise early adopters, and operators who want to be at the frontier of autonomous swarm deployment. If that is you, let's talk.

// Contact

Build With Us. Fly With Us.

Name is required.
A valid email is required so we can reply.
Message is required.
Platform
Company
SkyFlock UAViation Pvt. Ltd.
Incubator
IIT Ropar
Location
Haryana, India