MBZIRC 2020: Team Tartans

Published: by

MBZIRC 2020: Team Tartans

The Mohamed Bin Zayed International Robotics Competition seeks to exhibit the current state-of-the-art in real-world autonomous multi-robot systems. Organized by Khalifa University, the competition is held every two years with $5M in prize money and team sponsorship. Teams from around the world, including top robotics research universities and companies, compete over a 3-day period and are judged by a panel including robotics researchers. From the MBZIRC website:

Robotics has the potential to have an impact that is as transformative as the internet, with robotics technology poised to fuel a broad range of next-generation products and applications in a diverse array of fields. Robotic competitions in the past few decades have been a catalyst that has accelerated the rate of technological advancements in the field of robotics and autonomous systems.

MBZIRC aims to provide an ambitious, science-based, and technologically demanding set of challenges in Robotics, open to a large number of international teams. It is intended to demonstrate the current state of the art in robotics in terms of scientific and technological accomplishments, and to inspire the future of robotics.

Similar to other major competitions, the MBZIRC aims to provide an environment that harbours innovation and technical excellence, while encouraging spectacular performance with robotics technologies.

Challenge Description

Challenge 1 aims to tackle the problem of airspace safety with both static and dynamic actors. The task includes autonomously popping balloons placed randomly throughout an arena and catching a ball hanging from a drone that is flying overhead. This must be done with multiple UAVs.

Challenge 2 tackles the problem of autonomous construction with aerial and ground robots. UAVs must pick blocks up from a pattern and place them on an elevated structure. The UGV must pick up blocks and place them on a designated placement pad.

Challenge 3 focuses on firefighting with aerial and ground robots. Fire blankets must be placed on simulated fires on the ground and water must be injected into real fires on the outside of a three-story building. The Grand Challenge comprises of tasks from all three challenges performed together, with a limited number of robots operating at any given time.

Details of all challenges can be found here.

Results

Challenge 1: 8th place, 5th highest autonomous-mode score

Challenge 2: 4th place

Challenge 3: 16th place

Grand Challenge: 7th place (5th for Challenge 1 tasks, 9th for Challenge 2 tasks, and 4th for Challenge 3 tasks)

Key achievements
  • Challenge 1: autonomously popped all balloons and successfully servoed towards ball

  • Challenge 2: one of only five teams to autonomously pick up and place a block with a UAV

  • Challenge 2: successfully accomplished six autonomous UAV block pickups in 25 minute period

  • Challenge 3: autonomously shot most water onto outdoor fire with a UAV; only team to do so on a target which had the additional wind disturbance

  • Challenge 3: only team to autonomously fly inside building (through 2nd story window)

The Team

Team Tartans
Team Tartans traveling team in the grandstands of the MBZIRC 2020 event

This project is a collaboration between the AirLab and the Intelligent Autonomous Manipulation (IAM) Lab and includes students from multiple programs.

PIs

Sebastian Scherer (AirLab)

Oliver Kroemer (IAM)

Students

Anish Bhattacharya (student lead)

Akshit Gandhi

Shubham Garg

Ganesh Iyer

Noah LaFerriere

Parv Parkhiya

Lukas Merkle

Andrew Saba

Rohan Tiwari

Stanley Winata

Karun Warrior

Special Thanks

Yaoyu Hu

Cameron Kisailus

Valmiki Kothare

Jay Maier

Lorenz Stangier

Wenshan Wang

Kevin Zhang

In the media

Article in CMU News

Article in the The Engineer

Contact

Anish Bhattacharya, anishb [at] cmu [dot] edu

Sebastian Scherer, basti [at] andrew [dot] cmu [dot] edu

Oliver Kroemer, okroemer [at] andrew [dot] cmu [dot] edu

Project Overview

Latest Research

SubT-MRS: Pushing SLAM Towards All-weather Environments
SubT-MRS: Pushing SLAM Towards All-weather Environments

Simultaneous localization and mapping (SLAM) is a fundamental task for numerous applications such...

TartanDrive 2.0: More Modalities and Better Infrastructure to Further Self-Supervised Learning Research in Off-Road Driving Tasks
TartanDrive 2.0: More Modalities and Better Infrastructure to Further Self-Supervised Learning Research in Off-Road Driving Tasks

TartanAviation: Image, Speech, and Trajectory Datasets for Terminal Airspace Operations
TartanAviation: Image, Speech, and Trajectory Datasets for Terminal Airspace Operations

We introduce TartanAviation, an open-source multi-modal dataset focused on terminal-area airspace...