Sat 12-Oct |
Sun 13-Oct |
Mon 14-Oct |
Tue 15-Oct |
Wed 16-Oct |
Thu 17-Oct |
Fri 18-Oct |
|||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
AI Olympics With RealAIGym | Welcome Ceremony | Tidylng Up | |||||||||||||
The Earth Rovers Challenge | |||||||||||||||
Robotic Construction Challenge | |||||||||||||||
euROBIN Manipulation | |||||||||||||||
Underwater Robotics Challenges | |||||||||||||||
Automatic Data Annotation Challenge* | |||||||||||||||
F1 Tenth Autonomous Racing Competition | |||||||||||||||
IEEE/RAS Quadruped Robot Challenge (QRC) | |||||||||||||||
Falcon Monitoring System** |
This competition focuses on the physical or athletic intelligence of robots.
To facilitate reproducibility and to stimulate algorithmic advancements, the AI Olympics competition is being held at IROS 2024 based on the RealAIGym project. The challenge will involve two stages:
People from all communities (AI, ML, RL, Optimal Control, Heuristics, or any combination of these approaches)are invited to try this competition and submit their best efforts (in form of 2-4 page paper with accompanying code) in achieving very standard dynamic task (i.e. swingup and upright stabilization) on a canonical underactuated system known as acrobot or pendubot (double pendulum with single actuator).
This competition focuses on remote robot navigation.
with the robotics community now exploring challenging open-world and open-vocabulary mobility scenarios. The competition seeks to conduct globally distributed, in-the-wild navigation evaluations on an unprecedented scale, releasing a significant real-world navigation dataset from over 10 cities (>1k hours). By deploying a large fleet of outdoor navigation robots across multiple cities, we aim to assess the performance of state-of-the-art open-world autonomous navigation models in true open-world settings and compare them to human teleoperated performance. The competition will take place in at least four cities (e.g., Abu Dhabi, Madrid, Taipei, Stockholm), requiring participants to deploy their navigation policies in realistic, goal-oriented scenarios, potentially without known maps, to test the models' robustness, generalization, and safety.
This competition focuses on robot manipulation.
When envisioning the future of the construction industry, robots have already made significant progress, offering benefits such as increased efficiency, enhanced safety, and improved building quality while reducing costs.
Intelligent robots, with advanced perception, decision-making, and execution capabilities, will become crucial in enabling more adaptive and efficient construction processes. To promote embodied intelligence in construction, we are launching the inaugural Robotic Construction Challenge. Participants will use building blocks to recreate the leaning shape of the Capital Gate in Abu Dhabi, employing single or dual robotic arms for identification, location, grasping, and stacking tasks using visual recognition and control algorithms. We will provide a workbench, standard building blocks, and a simulator.
This competition focuses on robot manipulation
This competition is an initiative aimed at showcasing the forefront of robot manipulation capabilities. The challenge invites global research institutions to enhance manipulation techniques through an electronic task board featuring tasks like
Teams demonstrating quick and automated solutions will present live at IROS, with an expert panel assessing their adaptability to new scenarios, acknowledging the emerging field of skill generalization.
This competition focuses on Navigation.
It is designed to inspire and engage high school students from the MENA region and developing countries in Africa, teams comprising five students aged 14-17 will participate in tasks focused on balancing with additional payloads and delivering objects over specified distances.
Task 1: the robot should be able to balance its weight with additional payload of 500 grams (500 ml bottle water) on a flat horizontal surface while remaining at rest for a period up to 3 minutes.
Taks 2: In this task, the robot should balance itself on an inclined ramp of 15º with the bottle.
Task 3: In this task, the robot should be able to deliver a payload of 500 grams (bottle of water) for 1 meter on a horizontal surface. The task should be completed within a period of 100 seconds.
Task 4: In this task, the robot should be able to deliver a payload of 500 grams (bottle of water) for 1 meter on an inclined surface of 15º. The task should be completed within a period of 100 seconds.
This competition focuses on underwater robotics.
The Underwater Robotics Challenges, organized annually in the MENA Region by the Underwater Robotics Research Center, serve as a platform for innovation and problem-solving in the fields of underwater robotics and artificial intelligence. These challenges aim to address various issues, including water and sea pollution monitoring, marine tourism, and the conservation of marine life such as fish and coral reefs. Held at local, regional, and international levels, the challenges provide opportunities for participants to showcase their designs and ideas, fostering collaboration and advancement in marine technology and robotics.
This competition focuses on Perception and Sensing.
The motivation behind this competition proposal is to facilitate the automation of annotations in the field of autonomous driving. Training deep-learning models for autonomous vehicles requires massive, annotated datasets for tasks such as detection, segmentation, tracking, and more. Automating the annotation process can significantly accelerate research and field evaluations. The competition aims to design an automated annotation pipeline for the ROAD dataset, generating annotations in the existing format along with textual descriptions. The ROAD family datasets include videos from the UK, the US, and the UAE. Automating annotations will speed up the utilization of new data and eliminate the need for manual annotations. Textual descriptions will provide deeper insights and can be used to train Vision Language Models (VLM) and hybrid models for road event detection
This competition focuses on Navigation.
The teams participating in the F1TENTH Grand Prix at IROS 2024 will build a 1:10 scaled autonomous race car according to a given specification and write software for it to fulfill the objectives for the competition: race head-to-head, Don’t crash, and minimize lap time. Each team will bring its own physical F1TENTH car with its own written software. The organizers provide the race setup (rules, submissions, guidelines), the track, and related infrastructure, as well as organize the race itself.
This competition focuses on Navigation.
The IEEE RAS Quadruped Robot Challenge (QRC) initially conducted at ICRA-2023 in London, England was the first in a series of challenges for autonomous and remotely-controlled quadruped robots using standard terrains, obstacles, and tasks to quantitatively measure and compare capabilities. These challenges inspire innovation in legged robots and will evolve to include embedded dexterity tasks for onboard manipulators and embedded mapping fiducials to evaluate 2D/3D maps generated within complex environments. These challenges can also be expanded to include humanoid robots with the simple addition of an overhead safety belay on a rolling gantry to catch and reset falling robots.
The multi-award winning Abu Dhabi National Exhibition Centre (ADNEC) is the largest exhibition venue in the Middle East, fully interconnected within the hearth of Abu Dhabi. Featuring a unique architectural design, ADNEC offers one of the world's most accessible and flexible event spaces.