2010 Virtual Manufacturing Automation Competition

Tue, 10/13/2009 - 00:34 — admin

The competition is in conjunction with IEEE ICRA
New:

Videos and Results from ICRA 2010
Results from ICRA 2010

Video shows ICRA 2010 runs and working on a real KUKA robot cell.

Video shows a multi-robot simulation environment for mixed palletizing.

VMAC Mobility Competition.




About the competition:

Overview

Entries are sought for the 2010 Virtual Manufacturing Automation Competition (VMAC) to be held in conjunction with the IEEE International Conference on Robotics and Automation (ICRA), Anchorage, Alaska taking place on May 3-8, 2010. Participation in the competition will consist of a software demonstration of a team of robots solving one or more of the pre-defined challenge tasks in the USARSim simulation environment. This demonstration may be performed remotely or on-site in the simulated environment set up for the competition.
The goal of the challenge is to provide an industrially relevant scenario and performance benchmarks to assess technologies in the areas of robot navigation in dynamic unstructured environments, mixed palletizing operations, and mobile manipulation. The competition’s open source policy is designed to encourage collaboration and the dissemination of ideas and algorithms.

The VMA competition is being conducted under the auspices of the IEEE Technical Committee on Performance Evaluation and Benchmarking of Robotic and Automation Systems (TC-PEBRAS).

Background

There exists a desire to limit the manual handling of goods and the necessity for people to work in difficult environments while reducing errors and increasing throughput. Automated Guided Vehicles (AGVs), forklifts, and robotic arms represent integral components of today’s manufacturing processes. They are widely used on factory floors for creating pallets of mixed goods for transport, intra-factory transport of goods between conveyors and assembly sections, parts and frame movements, and truck-trailer loading/unloading. Automating these systems to operate in unstructured environments presents an exciting area of current research in robotics and automation. Unfortunately, the traditional entry barrier into this research area is quite high. Researchers need an extensive physical environment, robotic hardware, and knowledge in research areas ranging from mobility and mapping to behavior generation and scheduling. An accepted approach to lowering this entry barrier is through the use of simulation systems and open source software. This year’s competition will build on the successful 2009 VMAC and include both mixed palletizing operations and material transport.

Competition Format

The competition will include several “elemental” tests as well as a full scenario. Teams are encouraged to participate in as many events as possible, with winners in each of the individual events as well as an overall winner being announced. Since this is a simulated event, perfect ground truth is available on such items as vehicle locations, package locations, package types, etc. It is desired that teams utilize as little of this information as possible, however, the organizers realize that most teams will require some help from ground truth. Teams must disclose their ground truth needs in their Team Description Paper (described below).

Elemental Tests

The elemental tests are designed to address competencies in individual skills. This year’s tests will address mixed palletizing and vehicle mobility.

  1. Mixed Palletizing Elemental Test

    The objective of this test is to autonomously fill orders that a distribution center receives. The teams will have up to 3 robotic arms available, and each arm will work simultaneously on up to 3 pallets of materials. This test will consist of a circular conveyor (speed controlled by the team) that contains a statistical distribution of packages. Teams will receive their packing lists and must assemble a stable configuration of the list’s contents on the pallets as rapidly as possible. The competition will consist of several rounds, with the number of package classes increasing with each round. Teams are free to use any USARSim supported robotic arm, and may hard mount the arm or mount it on a mobile platform. A sample world will be provided in the near future for experimentation with this event. Teams should include a description of their anticipated set-up in their Team Description Paper as described below. Teams should also include any underlying simplifying assumptions that they will make and any ground truth that they will require. Ground truth that could be provided includes both local information such as an individual package class and location, and global information such as all package’s classes and locations.

  2. Mobility Elemental Test

    The objective of this test is to autonomously deliver completed pallets throughout an unstructured factory environment. Teams will have up to 3 robotic platforms available for delivering the pallets. When a pallet becomes available, the team must dock a robot with a loading conveyor to pick-up the pallet, and then deliver it to one of several locations. The delivery environment will be maze-like and may contain dynamics and unexpected closed pathways. This competition will also consist of several rounds with the complexity of the environment increasing with each round. Teams should include a description of their anticipated set-up in their Team Description Paper as described below. Teams should also include any underlying simplifying assumptions that they will make and any ground truth that they will require. Ground truth that could be provided includes a priori maps of the environment, vehicle locations, and package delivery points.

  3. Scenario

    The full scenario will combine the two elemental tests into a single event. Teams will prepare pallets for transport and then move them throughout the factory environment to their destination.

  4. BONUS

    As an added bonus, a Kuka robotic arm with a suction end-effector will be available for use during the competition. The robot will have an identical API as the virtual Kuka robotic arm. Code that successfully completes the mixed palletizing elemental test will be given the opportunity to run on the real hardware.

What’s Next?

Potential participants should submit a team description paper (TDP) of their intended entry to roboSim@nist.gov. The TDP should contain which challenge events your team will participate in, a high-level description of the algorithms that you will employ, requirements that you will place on the simulation system (i.e. what robots, sensors, and infrastructure you expect to need), and references to the team’s relevant work in the area. If the number of teams wishing to participate exceeds the hosting capabilities of the competition, then the VMA technical committee will use these TDPs to select participants.

Resources