Developments in digitisation: STEINERT

Developments in digitisation: STEINERT

STEINERT’s Johann Hefner outlines the role artificial intelligence and digitalisation will play in Australia’s material recovery future.

To effectively keep resources in the value-chain and out of landfill, waste operators around the global are looking towards innovative Industry 4.0 approaches such as digitisation, artificial intelligence and robotic technologies.

Described as the fourth industrial revolution, Industry 4.0 refers to the current trend of continued technological improvements and digitisation.

To help industry better understand these technologies, magnetic and sensor-based sorting solutions leader STEINERT ran an online seminar in late February – discussing its experience and developments in the digitisation of the waste and recycling industry.

The seminar explored a range of concepts and technologies, with STEINERT UniSort Managing Director Hendrik Beel using practical examples to demonstrate how STEINERT works collaboratively with its customers to develop highly efficient sorting facilities.

In 2018, for example, STEINERT delivered a fully automated plant at one of the world’s biggest material recovery facilities in Germany.

The project was conceptualised around STEINERT’s new facilities framework, which calls for high levels of automisation to support improved recycling rates, while meeting demand for final product purity.

According to Johann Hefner, STEINERT Australia Resource Recovery Manager, the plant processes lightweight packaging waste from roughly 6.4 million people, with a capacity of 200,000 tonnes per annum.

STEINERT’s design and technology ensures an end-product with the highest purity levels available to industry, allowing its customer to replace raw materials from primary sources that are running low.

The plant features 58 near-infrared sorting systems, four Unibot robots, five non-ferrous metal separators and four overhead suspension magnets.

“The plant is fully automated, so no hand pickers are required, and all near-infrared sorters are interconnected to achieve the best purities. Additionally, as the entire system is automated and monitored, the facility can be operated and maintained with minimal staff,” Hefner explains.

DIGITISATION AND DATA

While data is seen as the key to previously irresolvable challenges in the recycling industry, Hefner says it is an often-under-utilised resource of companies operating in this space.

“There are calls for larger, more complex sorting facilities as a result of greater financial pressure, stricter requirements based on recycling rates and the materials in material flows comprising multiple layers,” he says.

“As such, facilities which are well utilised, easy to monitor and quick to maintain are absolutely essential for smooth-running and economically viable sorting processes.”

Hefner adds that STEINERT believes artificial intelligence will play a major role in the future of raw material recovery.

“A future that can offer unimagined opportunities in the fight against the scarcity of resources based on huge amounts of digitally obtained data and the mechanical evaluation of this data is central to STEINERT’s business model,” he says.

Hefner explains that these solutions build on STEINERT’s long-history in waste and recycling technology, specifically noting its UniSort PR.

“In 2011, the first Hyperspectral Imaging (HSI) camera replaced the point-to-point scanners previously used as standard in the STEINERT UniSort PR, paving the way for data-based sorting solutions,” he says.

STEINERT’s HSI camera has a spatial and spectral resolution of 28,000,000 pixels, a 5000-fold improvement, and forms the basis of the company’s networked solutions – thereby providing new opportunities in the fight against the scarcity of resources.

“With the accumulated experience from four UniSort generations, a system was developed to fully process this data,” Hefner says.

All the necessary components are built into the machines of the EVO 5.0 generation, with customers benefiting from shorter commissioning times, easier facility monitoring, robot-based quality control and spectral databases with AI deep learning that helps sort previously inseparable materials.

“This enables the highest sorting performance for every machine globally all the time, as the database and calibrated system is identical for every machine. Additionally, the database is fuelled by an AI algorithm, so we continue to learn and improve the sorting efficiency and can continuously update the customer’s software.”

Hefner adds that STEINERT’s intelligent plant commissioning capabilities facilitate the commissioning of all sizes of plant and synchronisation at the touch of a button.

“This cuts commissioning times while also offering optimised staff management and enabling the commissioning of sorting facilities with more than 50 near-infrared sorters,” he says.

“By calibrating machines using a central spectral database, the entire system is adapted to the local circumstances at all times and is stable in the face of environmental factors.”

Similarly, STEINERT’s intelligent remote monitoring cuts the time it takes to respond to problems thanks to mobile status monitoring and opportunities for service experts to intervene.

“This provides overall plant performance improvement, as well as shorter downtimes and overall operational flexibility,” Hefner says.

“Our intelligent remote update also allows improvements or totally new functions to be deployed remotely and equips sorting facilities for future requirements without staff having to be present on site.”

On a more granular level, STEINERT’s intelligent object identified system uses AI to sort materials that could not previously be separated and for solving complicated sorting tasks, thereby guaranteeing constant and improved sorting performance.

“This allows for faster adaption to changes in material compositions and requirements, with the AI learning more stable visual differences,” Hefner says.

The AI object identifier adds image information to the sorting decision so operators can detect and sort individual item, for example cleaning detergent bottles that are of a target recyclable plastic but could contain toxic substances.

“Our robot-based intelligent quality control then checks for automated assessment of material qualities for processing, using for example the UniSort Unibot,” Hefner says.

“This enables accurate and flexible pricing as well as a response to legal and customer-specific requirements.”

For more information click here. 

Related stories:

X