An illustration of the washroom environment to aid proposal writing. Please note that the actual washroom that will be used in the competition may differ in some aspects.
The challenge requires a robot to perform certain types of janitorial tasks in a confined area of approximately 3m × 3m. The wash room will be covered by a ceiling or net at a height of about 3m. Since this is an indoor application, availability of GPS is not assured. Therefore, the solution should not use GPS. The wash basin counter of roughly 120cm x 60cm will be about 1 meter from the ground, and will be fixed to the wall. The wash basin will be roughly 50cm x 40cm, and will be about 10cm deep. It will have a smooth shape and a distinct colour. However, it will be placed towards the left or right end of the counter and will not be centred. The washroom will have a single door and may have a couple of windows. The height of the entry door would be between 2 m and 2.5 m and width would be in the 80cm - 100 cm range. The walls may have some texture or designs painted on them. The wash room will have a commode. It may also have an additional trash can for use by the customers. However, it will be distinguishable from the dustbin that the robot will use. The height of the dustbin (used by the robot) would be approximately 50 cm and it will be able to hold around 30 litres of trash. The robot will be given an opportunity to scout the area inside and outside of the washroom (either manually or autonomously) for a limited time. A pneumatic cylinder or air actuator can be used with a maximum pressure limit of 600 KPA. The specific tasks that the robot should perform are listed below.
The main tasks to be performed by the robot are the following:
Task 1: Entry through the door – The robot will start from a location outside of the washroom. The exact location will not be specified ahead of time. It can be different for each team and even for the same team, it can vary in each run. The robot needs to detect the door and enter through it.
Task 2:Picking up trash from the floor – The robot needs to detect trash on the floor and pick it up. The trash will consist of items such as crumpled tissue paper, paper cups, and cans. The exact nature of the trash will be specified more clearly in the coming weeks. However, it can be assumed that the type of trash is known and it can be identified by features such as colour, texture, shape, and hardness. The robot may need to apply different techniques for picking up the trash depending on the type of trash.
Task 3:Dropping trash in the dustbin – The location of the dustbin is not fixed and it can be anywhere in the wash room near the walls. Its presence, position and orientation can be identified by features such as size, shape, colour, and texture.
Task 4:Spraying a sanitizing solution on the counter – The robot can be assumed to possess sanitizing liquid onboard in a suitable container. It needs to be equipped with a proper mechanism to be able to reach the counter top and spray the sanitizer on it, as well as on the wash basin.
Task 5:Drying the wet countertop – The countertop would have been sprayed with sanitizing liquid and the robot should dry it.
Task 6:Cleaning the floor – The robot needs to spray sanitizing liquid on the floor on dirty spots and then clean them up. There will be dirt marks on the floor in random places (e.g., markings created by a marking pen), and the robot needs to remove the markings.
Task 7:Exiting through the door – The robot needs to find the door, exit the washroom and come back to the same location it started from.
Task 8: Drying the wash basin – The wash basin would have been sprayed with a sanitizing solution. The robot needs to dry it.
Task 9: Countertop cleaning with objects present – Do Task 5 with other objects on the counter (such as soap or tissue paper dispensers) without toppling them.
In general, the scoring will be based on the following considerations. A more detailed scoring system will be published in the coming weeks.
Number of successfully completed tasks (partial weightage may be given for partial completion of some tasks)
Degree of integration of map creation activity (done by the robot versus separately loaded) (fully integrated solution will be given extra points)
Minimization of the use of additional help for localization such as April tags or ArUcO markers.
Minimization of Collisions and any eventual damage to the walls, door frame, and other objects (large penalties for damage or collision with humans)
Time required to complete the tasks (faster the better for most tasks)
The amount of hardware required for the solution (total number of motors, GPUs, and their power) (lower the better).
Challenge versions of the tasks will get extra points.