octograbOctoGrab is an autonomous delivery drone designed as an extension of the SPHERES robots. It uses the existing VERTIGO project to achieve orientation and navigation inside the International Space Station and adds a new extension allowing to recognize and grab the objects to carry.
This project is solving the ZERO GEE Bee – Your Friendly Neighborhood Drone challenge. Description
The purpose of the project is to design a drone able to autonomously deliver objects inside the International Space Station and future large scaled spacecrafts in order to solve logistical and service needs of a human operated space flight. The two main differences between this project and the existing solutions are that the drone must be able to operate in: - a weightlessness and zero g environment; - in a confined environment close to human beings, which implies hard constraints related to the safety of the crew and also trying to be the less intrusive possible regarding their daily life and task aboard the station. It is also specified that the drone shall perform autonomous navigation inside within the station. We chose to extend its autonomy to all its payload related tasks: item localization inside the ISS and object grabbing/unloading.
After having studied the constraints and implications of the project, we determined that the SPHERES robots and related projects developed by the MIT answer many of our needs, particularly regarding the problematics of the positioning, orientation/propulsion and navigation of the drone. The SPHERES robots are operational aboard the ISS since 2006 and have already been used to prove several concepts to answer these problematics., Moreover, the SPHERES robot is designed as an evolutive platform, allowing to plug a custom extension (goggles, robotic arm, ...) to a dedicated extension slot. The MIT team have already designed some sub-projects that we will reuse directly or in a modified version. The VERTIGO goggles project aims at creating a device for three-dimensional mapping of a target and performing relative navigation. We plan to extend the capabilities of the device to allow mapping of the entire environment in the LOS of the drone in order to perform collision free navigation around all potential obstacles (floating objects, astronauts, ...). The Halo project is a structure providing six expansion ports allowing to significantly increase the amount of hardware usable simultaneously on a SPHERES satellite. We will create a similar structure to provide one extension slot on the front to attach an extended VERTIGO device for the navigation and two extension slots on the rear to attach a classic VERTIGO device and a grabbing mechanism to load the object to deliver.
To perform the navigation in the ISS we simply reuse the ecosystem of the SPHERES robots (fixed positioning beacons) extended to the entire station.
The new grabbing extension is made up of 4 articulated arms allowing to grab and hold its payload during the delivery. Their articulations allow to grab a vast range of objects of various dimension. The arms are moved thanks to a system of opposite wires and spring allowing to apply the exact pressure needed on the payload. To move large objects not fitting in the drone arms we also added a suction cup for allow fixing of large objects.
The autonomous object identification is performed with the classic VERTIGO device near the arms on the rear of the drone. It uses a (existing ?) database containing the signature shape of the all the objects inside the ISS. The location of the object to load and deliver can also be automatically determined using the existing manifests referencing the place where some objects can be found in the ISS.
License: GNU General Public License version 3.0 (GPL-3.0)
Source Code/Project URL: https://github.com/deshmal/OctoGrab