Robotics and Artificial Intelligence to solve the lack of human resources in recycling plants during a COVID-19-like crisis

by | Apr 20, 2020 | Artificial Intelligence, Blog, Circular economy, Deep learning, innovation, Machine Vision, Robotics | 0 comments

At the height of the COVID-19 crisis, waste management has become a priority for several countries, but keeping recycling plants at full capacity represents a challenge.
The collection and treatment of waste are essential services that must be guaranteed despite the health crisis. And the automation of tasks that entail risk or with low-added value is now more necessary than ever before.

Greater automation in the recycling industry does not just ensure the functioning of this service. It also represents an alternative for minimising worker contact with the waste. Furthermore, the installation of smart robots helps to guarantee compliance of social distancing measures that are likely to remain in place after the COVID-19 crisis.


Intelligent robotics improves occupational safety at recycling plants

Coronavirus has a direct impact on the recycling industry workforce, where occupational safety is essential. Further boosting automation in recycling plants by tapping into the advances made in Artificial Intelligence (AI), Machine Vision and robotics will help waste plant managers to minimise the risks operators face even more.

The benefits of automation through AI robots not only lie in better work environment safety, but end of line automation in waste treatment plants enables an increase in the efficiency of separating waste and in the collection of recoverable materials. Work is done without interruptions and with constant output.

Automation is nothing new in the recycling industry, but with the latest innovations in Artificial Intelligence and Vision, we’ve taken a big step forward in a 4.0 industry context. These innovations represent better productivity, but also effect human resources management. Transferring low added-value manual tasks from the hands of a worker to the arms of a smart robot such as Ecopick


The potential of Artificial Intelligence in a recycling plant

Progress in AI has revolutionised robotics. And has created innovative devices capable of displaying smart behaviour, even able to reproduce some human thinking capacities. These new robots open a range of opportunities for the industry in general and specifically for waste management and recycling.

Thanks to an AI system, these robots can select and classify the waste just as well as humans do, faster, and with a lower margin of error. We therefore have machines trained to do tasks that workers did until recently. They are now freed from these routine tasks with little added-value and are deployed to other, more productive and lower risk tasks. Furthermore, the use of smart robots improves recoverable product quality and reduces the amount of unsuitable material that reaches the end of the process.

Smart robots increase efficiency in the classification and triage of recoverable waste. AI used to be based on traditional vision, coming from deterministic contexts: the same objects, same position, same lighting conditions, etc. Now, through a computer vision system, information is captured through sensors that capture images analysed in real time through deep learning algorithms, to understand the context and take decisions. The same robot, thanks to the AI system, is able to interpret changes in the environment and adapt to them.


Ecopick, the smart robot that classifies recoverable waste as a manual picker would do

The Picvisa Ecopick is an example of this kind of robot, which uses the potential of Artificial Intelligence and Vision. Ecopick has a fully flexible robotic arm that can adjust its gripping system based on the type of object collected. Through an Artificial Vision system, it can scan the different products (PET and PEAD bottles, tetrabrick packs, paper/cardboard, etc.), identify them among the rest of the waste and separate them for recycling and recovery at a later stage.

Material detection is done using a visible imaging spectrometer (VIS-RGB) and, optionally, near-infrared spectroscopy (NIRS). It also includes the option of a remote connection for system updates and streamlining, expanding the library of objects to be recognised and enhancing precision over time.

Ecopick’s most innovative feature is that it works with a with an artificial vision system featuring AI (deep learning) algorithms, enabling it to recognise objects to separate on the conveyor belt. Specifically, in the first stage, it detects the shapes, colours and textures, and later, learns with human supervision to separate different waste, with the subsequent improvement in the quality of the material recovered.


What benefits does Ecopick provide recycling plants?

The benefits that Ecopick brings to the recycling industry are clear and assume even greater importance during the crisis we are facing due to the COVID-19 pandemic.
1.    Production stability. Ecopick performs 24/7.
2.    It enables manual pickers to be positioned in lower-risk roles with greater added value.
3.    It provides real time access to statistics on classified materials. 
4.    It facilitates and speeds up intervention for line settings, thanks to vision and data analysis.
5.    It has online customer service, without requiring technical visits. The robot has minimum maintenance that can be done remotely, which is extremely important during our current confinement.
6.    Our system adapts to existing installations.

Picvisa, with 15 years of experience in the recycling sector and a pioneer in computer vision application, has reached an agreement with Ferrovial to implement this robotics system in AI in one of its recycling plants.

For further information, download the Ecopick data sheet


Submit a Comment

Your email address will not be published. Required fields are marked *