Duckietown

Website Contact the school

Democratizing access to the science and technology of robot autonomy.

Originating at the Massachusetts Institute of Technology (MIT), Duckietown is a platform for delivering cutting-edge robotics and AI learning experiences tailored to all skill levels. We offer teaching resources to instructors, hands-on activities to learners, an accessible research platform to researchers, and a state-of-the-art ecosystem for professional training. We believe the world needs more talent to shape the next generations of robot autonomy, and we know that talent is you.

08/07/2026

A day of autonomous robotics, teamwork, and engineering challenges brought the Chilean Duckietown community together in Puerto Montt for DuckieDay 2026.

Hosted by Puerto Montt, students from six campuses across Chile came together to develop based solutions related to challenges arising in the national salmon industry.

Powered by Duckietown, teams developed autonomous Duckiebots capable of navigating independently, detecting sea lice using onboard computer vision, issuing an alert, and autonomously returning to their starting point.

We are proud to see Duckietown used as an enabling pedagogical tool for inspiring the next generation of engineers in robotics and physical AI.

Learn more about Duckie Day 2026: https://hubs.la/Q04sdb2P0

07/10/2026

Meet Vikram Radhakrishnan, Lecturer in applied data science and A.I. at The Hague University of Applied Sciences.

Vikram shares his experience bringing robotics and AI education into his classroom, and how hands-on platforms help students move from theory to real-world autonomous systems.

Read the full interview and discover his journey: https://hubs.la/Q04p87lH0

06/30/2026

What happens when robots can understand more than pixels? 👀🤖

Traditional autonomy architectures separate perception, planning, and control into distinct modules. Vision-Language Models (VLMs) instead combine visual perception with language-based reasoning.

This different approach enables interpretation of semantic information from a scene and incorporation of the findings into navigation decisions, alongside the existing autonomy stack.

The challenge is integrating these models into an autonomous navigation pipeline. Perception, planning, and control operate as interconnected modules, while VLMs contribute semantic reasoning. Defining how information flows between these components, and determining which decisions are delegated to each, is an active area of robotics research.

Check out how learners are exploring this approach in Duckietown: https://hubs.la/Q04n290H0

06/23/2026

Happy International Women in Engineering Day! 🥳

Did you know that (on average, in the US and EU) less than one in five engineering students, university faculty, and actual engineers in the workforce are women?

An important reason for this imbalance is the perception that STEM education fields are more for men, which biases the formative pipeline early on. In particular, robots and robotics are mostly associated with notions of power, speed, precision, and danger - attributes that are typically more appealing to men rather than women.

At Duckietown, we believe that all talents should have equal access to the best quality education. We support the next generation of robotics scientists and engineers by making it easier to learn, experiment, and build modern AI robotics systems in an environment that is colorful and cute, but at the same time as complex and articulated as robots can be.

Learn more about robotics and AI education with Duckietown: https://hubs.la/Q04mgqvl0

06/16/2026

What does obstacle avoidance in robotics mean? 🤖

Obstacle avoidance is the ability of an autonomous robot to detect objects in its environment and safely navigate around them.

There is no single best way to do it. Different robots, tasks and environments affect what the approach could be. Common strategies are:

📡 Reactive sensor-based methods

Robots use direct sensor measurements, such as time-of-flight sensors or ultrasonic sensors, to trigger immediate actions like stopping or turning.

🗺️ Path-planning approaches

Using cameras, lidars, or depth sensors, the robot builds a representation of the environment and computes collision-free trajectories.

🧠 ML-based methods

Machine learning approaches enable robots to learn obstacle detection and avoidance behaviors from data and experience, often through reinforcement learning or imitation learning techniques.

At Duckietown, both virtual and physical can be programmed and tested using all of these approaches.

For example, try out the ML-based obstacle detection and avoidance learning experience: https://hubs.la/Q04lxfMG0

06/05/2026

The 2026 IEEE International Conference on Robotics and Automation (ICRA 2026) is wrapping up in Vienna, Austria.

As the flagship conference of the IEEE Robotics and Automation Society, ICRA is a key venue for sharing advances in robotics, AI, robot automation, and autonomous systems. From research papers and technical workshops to industry exhibitions and robot competitions, the event highlights the ideas and technologies shaping the future of the field.

Conferences like ICRA play an important role in connecting research, education, and real-world deployment, helping accelerate innovation across academia and industry, and ultimately advancing the field of robotics for the benefit of humanity.

👉 Learn more about ICRA 2026: https://hubs.la/Q04kf33-0

06/04/2026

One of the main bottlenecks in robot autonomy development is the cost of performing real-world experiments.

Running experiments takes time and resources: setting up the system, running tests, collecting data, debugging, reproducing results, and validating behavior add up quickly.

Simulation helps streamline the process by enabling faster development and earlier validation before deploying to physical robots.

The Duckiematrix is a virtual environment built to support the development of autonomy workflows in Duckietown.

👉 Try it out starting here: https://hubs.la/Q04k2LV30

05/29/2026

How to help learners gain the skills needed to design, deploy, and maintain autonomous robots in the wild?

Preparing a hands-on robotics class is challenging and resource-intensive. Duckietown helps educators teach how the science and technology of autonomous systems work through learning activities built on professional software, hardware, and infrastructure used in robotics.

Whether you are launching a new robotics course, upgrading an existing curriculum, or building an embodied AI training program, Duckietown offers the tools, content, and support to get started: https://hubs.la/Q04jk_rB0

05/25/2026

Guillaume Gagné-Labelle, Gabriel Sasseville, and Nicolas Bosteels from the Mila Institute in Montreal, Canada, explore how computation delays affect reinforcement learning performance in Duckiematrix autonomous driving tasks.

Using Soft Actor-Critic models in Duckietown simulation, the team compares classical RL policies with action-conditioned Real-Time RL approaches under fixed and variable latency conditions, evaluating their impact on reward stability, policy performance, and episode length.

The project highlights how real-time constraints can influence autonomous driving behavior and policy robustness in embodied AI systems.

Learn more: https://hubs.la/Q04hNLqQ0

05/15/2026

We are often asked if Duckiebots can do line following, and get mixed reactions when we say it does lane following instead.

While line following can be achieved with a color sensor and on-off control, lane following uses vision, recursive estimation, and non-trivial mathematical assumptions, even with a structured environment.

Using camera measurements, Duckiebots detect lane boundaries, estimate their pose, and control steering to continuously drive at the center of the road, heading straight ahead.

Delving into the details of this fundamental autonomous driving behavior is a great learning experience. You can start here:https://hubs.la/Q04gMcJW0

Address

Boston, MA

Website

https://duckietown.com/

Alerts

Be the first to know and let us send you an email when Duckietown posts news and promotions. Your email address will not be used for any other purpose, and you can unsubscribe at any time.

Contact The School

Send a message to Duckietown:

Want your school to be the top-listed school/college in Boston?