Orchestrating the Future of Robotic Fleets.
We are building a secure protocol to allow autonomous vehicles, drones, and robots to safely coordinate decisions in real-time, solving critical labor shortages and fragmentation.
The ProblemAutonomous systems don’t talk to each other, leading to massive inefficiencies and security risks.
1
Vehicles only see with their own sensors, leading to congestion and inefficiency.
2
Fleets waste billions due to a critical lack of coordination and communication.
3
Safety is at risk without a universal, secure communication standard.
Echelon ProtocolA universal protocol for secure, verifiable, machine-to-machine coordination for intellitegent decision making.
• Shared dictionary-based schema for seamless communication.• Secure, verifiable messages to prevent cyberattacks.• Built on top of proven V2X and 5G technology.

What happens when machines talk.

Enhanced Autonomous Vehicle Navigation Lab

This enhanced simulation shows 8 vehicles navigating a complex 36×36 grid with multiple obstacle types. Gray squares are undiscovered walls (visible to you but unknown to cars), while red squares are discovered obstacles. In uncoordinated mode, each vehicle only knows what it has personally discovered. In coordinated mode, vehicles instantly share discoveries, dramatically reducing redundant exploration. Use the buttons below to start/stop the simulation, switch modes, randomize positions, and lock the layout. Make sure to lock the layout if you want to compare modes on the same obstacle/course setup. Currently, due to a bug, the program only runs normally when you lock the layout... sorry :(

Legend: Colored circles are cars (0-7), hollow circles are goals,gray squares are undiscovered walls,red squares are discovered obstacles, dotted lines show planned paths.

Algorithmic Obstacle Design: Obstacles are placed using a systematic approach with 3-4 horizontal walls (5×1), 3-4 vertical walls (1×5), 2-3 L-shapes, 2 diagonals, and small blocks. Each structure has built-in gaps and collision detection prevents overlaps or car entrapments.

Watch the dramatic difference: uncoordinated mode shows massive redundant discoveries as each car independently explores the same dead ends. Coordinated mode shows efficient exploration with instant knowledge sharing, leading to much shorter total paths and faster completion.

Our simulation shows the difference: when agents coordinate, congestion disappears and collisions are avoided.

We’re builders across AI, robotics, and business.

Jordan Marshall
Aerospace Engineering
Rohan Patel
AI For Business
Kavin Phaibiani
Applied Data Science
Alyssa Chan
Physics
Gavin Huang
Computer Science