The public Konnex testnet provides practical tasks for evaluating robotics and mapping models in realistic simulation and sensor setups.

Available task categories

Participants can access a set of scenarios designed to replicate industrial and research workflows for training and validation purposes.

• Drone navigation scenarios that require path planning and collision avoidance using onboard sensors.
• Robotic arm control tasks focused on manipulation, grasping, and precise object handling within constrained spaces.
• 3D mapping challenges involving reconstruction of environments from camera and lidar data streams.

Validation mechanism

Independent validators assess task outcomes by analysing camera feeds and motion traces to determine whether objectives were met.

The system records sensor logs and execution traces to support reproducible verification and to enable audit of validation results.

Payments and incentives

The Konnex team plans to introduce payments in stablecoins, executed only after successful result confirmation by validators.

This model aims to provide businesses with verifiable deliveries while creating monetisation paths for developers of effective AI models.

Use cases and implications

On a warehouse example, a robot receives an item relocation task while validators check camera and motion evidence to confirm completion.

Such workflows are intended to bridge the gap between experimental models and production-grade guarantees required by commercial operators.