Skip to main content
Back to challenges
Challenge

Google ADK & Devin Drone Logistics Safety Agent

Autonomous electric aerial mobility ventures need strict air-corridor compliance and telemetry verification before urban dispatch. You will build a multimodal agent system that evaluates flight routes, airspace restrictions, and weather safety metrics to approve routing decisions.

Business OperationsHosted by Vera
Sign in free to start
Challenge brief

What you are building

The core problem, expected build, and operating context for this challenge.

Implement a Google ADK agent with Gemini multimodal reasoning and Devin code validation to evaluate aerial mobility path constraints against aviation compliance standards.

Delivery guide

How work is evaluated

Evaluation

Verifies dynamic aerial clearance algorithms against simulated red-zone map breaches.

Datasets

Challenge datasets

Public datasets, plus the private datasets you have access to.

Evaluation rubric

How submissions are scored

These dimensions define what the evaluator checks and which criteria separate a passable run from a strong one.

Dimensions
2 scoring checks
Binary
2 pass or fail dimensions
Ordinal
0 scaled dimensions
Dimension 1

Red Zone Collision Check

Must fail routes intersecting restricted airport zones

Binary check

This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.

Dimension 2

Route Optimization Score

Efficiency of proposed alternate route distance • target: 0.95 • range: 0.8-1

Binary check

This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.

Learning goals

What you should walk away with

  • Build a Google ADK agent powered by Gemini Multimodal models for airspace map inspection

  • Leverage Cognition Devin automated agent workflows to generate and execute geospatial validation scripts

  • Enforce Digital Sky regulatory red-zone restrictions on flight trajectories

  • Generate real-time flight corridor approvals with cryptographic clearance tokens

Resources and assets

Reference links and supporting material

Dataset notes

Urban geospatial vector maps, simulated weather conditions, and Digital Sky restricted airspace coordinates.

Challenge setup

Verifies dynamic aerial clearance algorithms against simulated red-zone map breaches.

Challenge input

JSON with waypoints coordinates, battery status, and weather vectors

Google ADK

Builds Gemini-powered multimodal agents.

Cognition Devin

Generates dynamic python validation tools for geo-checks.

Evaluated output

JSON with clearance status, calculated risk index, and alternate route if rejected

Evaluation checks
  • Must fail routes intersecting restricted airport zones
  • Efficiency of proposed alternate route distance • target: 0.95 • range: 0.8-1
Configured evaluation evidence
  • Benchmark: AerialNav-Bench
  • Route Optimization Score target: 0.95
  • 1 public reference case
Configured runtime
  • Python execution harness
  • Python sandbox (unavailable on Versalist)

Summary shows the first task, up to three tools, and up to three evidence items. Open the technical recipe to view all configured items.

View technical recipe

Configured tools

Action Space
  • Google ADK · Required
  • Google · Optional
  • Google Gemini · Optional

Evaluation contract

  • Red Zone Collision Check · Weight 1
  • Route Optimization Score · Weight 1

Metrics

  • Route Optimization Score · Target 0.95

Benchmarks

  • AerialNav-Bench
  • robotics

Reference cases

  • 1 public reference case

Workflow tasks

  1. Task 1

    Input: JSON with waypoints coordinates, battery status, and weather vectors

    Output: JSON with clearance status, calculated risk index, and alternate route if rejected

    Execution harness: Python

Evaluation tools

  • Google ADK · Agent Framework
  • Cognition Devin · AI Engineering

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Google ADK & Devin Drone Logistics Safety Agent