More and more waste is being produced globally, and it ends up in waste management facilities. Waste Robotics is on a mission to develop cutting-edge waste management solutions that will dramatically increase the number of waste types that can be identified, recycled, and reused. With SpecimONE hyperspectral imaging platform, Waste Robotics can now gain ground-breaking insights into the chaotic world of material recovery and take sorting accuracy to the next level.

Waste Robotics aims to enable more precise, safer, and more profitable waste recycling on a global scale. They combine advanced waste-handling processes, computer vision, deep learning algorithms, hyperspectral imaging, and state-of-the-art robotic technologies.

–  We are a growing team of passionate engineers and operations professionals designing and deploying an AI-driven robotic solution for material recovery, says Ziad Akl-Chedid, the VP of Products at Waste Robotics.

The company is based in Canada, but its products waste waste-sorting solutions are used worldwide. As waste management is a very complex field, Waste Robotics is constantly looking for ways to improve waste management processes so that even more waste types can be separated.

Even though Waste Robotics’s overall sorting accuracy was on a good level, they saw that there was still room for improvement. The company started to look for a solution and decided to test whether Specim’s hyperspectral imaging cameras could help them take the sorting accuracy to the next level.

–  The hyperspectral information we get from Specim’s cameras is used alongside other cameras and profilers to feed our Sensor Fusion algorithm for better object detection and classification, Akl-Chedid explains.

Spectral-based material identification – in real-time

After implementing Specim’s hyperspectral cameras, Waste Robotics can now see more than just the 2D or 3D image. They can see the hyperspectral signature of those images, which, in some cases, is the only way to detect the material of an object even after it would visually be beyond identification.

–  Using the SpecimONE hyperspectral imaging platform for data processing and training allows us to have a real-time spectral-based material identification to augment our understanding of what is being presented to our robot for sorting, Akl-Chedid clarifies.

In addition, using SpecimONE and interpreting the hyperspectral data it produces is more straightforward than one might think.

– Another great thing about the hyperspectral imaging is that it is very visual. We can see the different colors on the image that represent different materials, Akl-Chedid clarifies.

A partnership built to innovate

Akl-Chedid has been pleased with the quality of Specim’s hyperspectral solutions and working with the company. He praises the reliability and responsiveness of Specim and their local representative.

In addition to providing training information, offering general support, and delivering equipment on the agreed schedule, Akl-Chedid values Specim’s understanding of their business area.

–  Our systems operate in extremely harsh conditions. The Specim team is aware of this and works with us to ensure our solution continues to perform optimally, Akl-Chedid says.

The two companies also share the same passion for innovation and pushing the limits of technology.

– At Waste Robotics, innovation is at the core of our day-to-day. We share this passion with our partners at Specim, he continues.

Akl-Chedid believes hyperspectral imaging will become increasingly widely used in the future thanks to its added accuracy.

– We see the need for hyperspectral imaging increasing as more material recovery facility operators discover the success stories we create together with Specim, Akl-Chedid concludes.

Related products:


Specim FX17