2026 Intern for Crowd-Sourced Mapping Algorithms
Core
Develop deep learning algorithms to convert crowdsourced trajectory data into accurate road-level networks and lane-level topology for mapping.
Role type
Intern, machine-learning engineer (mapping)
Builds
Server-side crowdsourced mapping systems
Domain
Autonomous driving, mapping, computer vision
Deliverable
production ML models
Required skills
Python, deep learning frameworks (TensorFlow, PyTorch), computer vision, graph-based models
Preferred skills
Surveying/mapping background
Responsibilities
Prepare and label datasets for training robust models; Design and fine-tune neural networks to convert crowdsourced trajectory data into accurate road-level networks; Create algorithms for generating lane-level topology in complex intersections and roundabouts; Develop strategies to ensure model resilience against noisy or incorrect data.
