DirtAI. “Research in Artificial Vision techniques for the Automation of Industrial Processes through the early identification of dirt.
Program: support grants for Innovative Business Groups to improve the competitiveness of small and medium-sized companies and the call for applications for the year 2021, within the framework of the Recovery, Transformation and Resilience Plan.
Referencia: AEI-010500-2021b-172
Date of RESOLUTION: 03/18/2022
Execution date: 18/11/2021-21/08/2022
In recent years, we have been observing a progressive automation of particularly tedious or repetitive manual tasks with low added value. Cleaning tasks, for these reasons, have been one of the first to incorporate mechanization and automation, and there have been a wide variety of commercial solutions at the domestic level for more than a decade.
Industrial cleaning systems are no strangers to this phenomenon, with many autonomous systems having appeared in recent years. However, all currently available autonomous systems base their operation on the repetition of a pre-established pattern, without previously assessing the level of soiling or having any feedback as to whether the cleaning has been performed correctly. In addition, in this type of environment there are a series of boundary conditions that can dynamically modify the organization and/or development of cleaning tasks, some of which are mentioned below as examples, as well as their effect on the planning of cleaning tasks.
– Limitation of the effective time available for cleaning → need to prioritize cleaning of some areas over others.
– Different levels of soiling in different areas → priority for cleaning the dirtiest areas.
– “Stubborn” dirt → Need to devote more effort in a particular area.
Thus, the main objective of the DirtAI project is the research on artificial intelligence techniques for the detection, classification and automatic positioning of fouling levels in industrial environments.
The following specific objectives are also proposed:
– Identify all types of soiling together and make a classification of levels and types of soiling.
– Identify the set of sensors required for the detection of dirt levels and types of dirt.
– Identify the appropriate measurement procedure to distinguish dirt from what is not.
– Establish optimal working conditions (light, humidity…).
– Avoid fixed infrastructure.
– INNOVATIVE LOGISTICS ASSOCIATION OF ARAGON (ALIA)
– SERVICE AND MACHINERY ARAGONESES S.A.
– LAFTUR SAL
– TECHNOLOGICAL INNOVATION CENTER FOR LOGISTICS AND ROAD FREIGHT TRANSPORT “CITET”.
– TECHNOLOGICAL INSTITUTE OF ARAGON.
TOTAL PROJECT BUDGET: 89.933,00€.
TOTAL BUDGET ITAINNOVA: 30.000,00 €.
TOTAL PROJECT FINANCING: 65.946,00€.
TOTAL FINANCING ITANNOVA: 19.500,00€.
This project is funded by the Ministry of Industry, Trade and Tourism of the Government of Spain, under the support program for innovative business clusters and by the European Union “NextGenerationEU” PRTR, being Project Reference AEI-010500-2021b-172.
