ML Intern: World Models for Autonomous Construction

Bedrock Robotics Inc; San Francisco (CA); On-site; USD 55,104,000 - 82,656,000; Full time

Job summary

Bedrock Robotics Inc is seeking an intern to help advance autonomous construction technology. You will train world models on real fleet data from autonomous machines in the field and run architecture explorations to identify robust approaches.

You’ll collaborate with the Behavior and Controls teams to translate model outputs into planning and control signals. As part of a team with veterans from Waymo, Segment, and Uber Freight, you will evaluate models, compare architectures, and build

Qualifications

• Pursuing BS, MS, or PhD in CS/ML/robotics or equivalent experience.
• Strong Python and PyTorch experience.
• Ability to translate research papers into working implementations.
• Experience with real-world, messy data.
• Good experimental judgment.

Responsibilities

• Train and evaluate world models on real fleet data from autonomous machines in the field.
• Run architecture explorations and ablations, and justify what works.
• Build evaluation harnesses and metrics to assess model improvements.
• Collaborate with behavior and controls teams to connect model outputs to planning and control.
• Present findings and provide a solid foundation for team work.

Skills

Python

PyTorch

Reading papers

Real-world data

Experimental judgment

Education

BS/MS/PhD in CS/ML/Robotics or equivalent experience

Job description

Bedrock Robotics Inc is seeking an intern to help advance autonomous construction technology. You will train world models on real fleet data from autonomous machines in the field and run architecture explorations to identify robust approaches.

You’ll collaborate with the Behavior and Controls teams to translate model outputs into planning and control signals. As part of a team with veterans from Waymo, Segment, and Uber Freight, you will evaluate models, compare architectures, and build