Robotics Intern - Large Behavior Models at Toyota Research Institute

Cambridge, Massachusetts, United States

Apply Now

Not SpecifiedCompensation

InternshipExperience Level

InternshipJob Type

UnknownVisa

Robotics, Automotive, AI ResearchIndustries

Requirements

Currently pursuing a degree (Ph.D., M.S.) in Robotics, Computer Science, Mechanical Engineering, or a related field
Publication record at top-tier robotics/ML conferences (RSS, CoRL, ICRA, NeurIPS, ICML, ICLR, CVPR, ICCV)
Hardware experience, especially toward deploying learned policies on real robotic systems (strongly preferred)
Experience with machine learning and familiarity with large datasets and models
Strong software development skills in Python (C++ experience very helpful, but not strictly required)
A “make it happen” attitude and comfort with fast prototyping and running informative experiments
A passion for the work
Comfort working with both existing large static datasets and a growing and dynamic corpus of robot data

Responsibilities

Create working code prototypes
Interact frequently with team members
Run experiments with both simulated and real (physical) robots
Participate in publishing the work to peer-reviewed venues
Work on research thrusts such as: data-efficient and general algorithms for learning robust policies leveraging multiple sensing modalities (proprioception, images, force, dense tactile sensing); scaling learning approaches to large-scale models trained on diverse sources of data (web-scale text, images, video); quick and efficient improvement of learned policies; developing and deploying learned policies on complex mobile manipulator embodiments (e.g., humanoid robots)
Focus areas may include: Vision-Language-Action (VLA) models for mobile manipulation; dynamic whole-body manipulation on humanoids; cross-embodiment transfer; universal (UMI-style or ego-centric) data collection methods; haptic/tactile-inclusive VLA models; integration of VLA policy methods with model-based robotics methods; large-scale synthetic data generation and sim-to-real transfer; post-training for continual learning

Skills

Robotics

Machine Learning

Generative AI

Foundation Models

Robot Manipulation

Sensor Data Processing

Simulation

Dexterous Tasks

Embodied AI

Large-Scale Datasets

Toyota Research Institute

Research in mobility, safety, and automation

About Toyota Research Institute

Toyota Research Institute focuses on improving mobility through research and development in the automotive and technology sectors. The company works on enhancing safety, automated driving, robotics, materials science, and machine learning. Their products include advanced safety features and automated driving systems that aim to make driving safer and more efficient. Unlike many competitors, TRI emphasizes a research-driven approach, collaborating with various partners and licensing their innovations to enhance Toyota's offerings and maintain a competitive edge. The goal of TRI is to advance mobility solutions that improve quality of life and support the transition to zero-emissions transportation.

Los Altos, CaliforniaHeadquarters

2016Year Founded

$100MTotal Funding

SEEDCompany Stage

Robotics & Automation, Automotive & Transportation, AI & Machine LearningIndustries

201-500Employees

Benefits

Highly competitive benefits package

Robust programs to support the wellbeing, happiness, and health of our people and their families.

401(k) plan including matching and annual profit sharing along with total vacation and holidays totaling 38 days per year.

Subsidized commuter benefits and generous employee and vehicle allowances

2 paid days per year to participate in volunteer activities.

Maternity Leave program with 10 paid weeks plus baby bonding leave and Milk Stork for traveling moms

Baby Bonding Leave–an additional 16 paid weeks–to all new parents, including those who choose to adopt.

Back-up child and adult / elder care programs to help everyone thrive

WellBeats Virtual Fitness Trainer.

Risks

Public skepticism may undermine AI-assisted driving safety initiatives.

Boston Dynamics' high costs could impact TRI's humanoid robot project returns.

Complexity of real-world environments may delay TRI's robot skill deployment.

Differentiation

TRI focuses on AI to enhance human life and mobility.

Partnerships with Boston Dynamics and Stanford showcase TRI's collaborative innovation.

TRI's Driving Sensei concept integrates AI to improve driver skills and engagement.

Upsides

Generative AI allows TRI's robots to learn complex tasks without new coding.

TRI's AI-driven material synthesis could revolutionize EV battery development.

Autonomous tandem drifting collaboration with Stanford pushes vehicle dynamics boundaries.

Land your dream remote job 3x faster with AI

Try Jobo Free