Tech Intelligence Bulletin (HG Insights) – itelligence has won an SAP Innovation Award in the ‘Process Innovator’ category with a project involving the use of drones and image recognition for the purposes of environmental protection.

These awards honor companies that operate successfully in the digital economy making use of SAP technologies in innovative ways. Norbert Rotter, CEO of itelligence AG, said, “I’m delighted that SAP is honoring itelligence AG as a ‘Process Innovator’ for an intelligent environmental protection project in this high-caliber worldwide competition. Being recognized as a good example of the smart use of digitization makes me very proud and will serve as a special incentive for us to warrant the trust placed in us in the future as well.”

SAP Innovation Awards selection

The SAP Innovation Awards team selected itelligence as one of its 2019 Process Innovator Award winners for an application of SAP technology that automated the detection of invasive species using drone images. By exploiting new technologies such as drones, image recognition, detection using neural networks, and deep learning, the itelligence team was able to detect the highly toxic and invasive species, giant hogweed, at rates better and much faster than those possible with human resources.

“Drones photograph the landscape and a previously trained and optimized algorithm processes the image data automatically and detects the weeds,” said Mark Albrecht, Global Head of Innovation Technologies at itelligence AG. “This involves products from the SAP innovation portfolio among others.”

The itelligence environmental protection project, which was conducted by itelligence Denmark, itelligence Innovation Technologies, and NTT DATA Japan, uses drones to loom for giant hogweed. The sap from these leaves causes photosensitivity in people’s skin, potentially resulting in serious skin inflammation. As this is a highly invasive species, land and forest owners have a duty to uproot the plants in order to eradicate them.

Even though the weed is quite large, it is laborious and time-consuming to scan swathes of land for it by eye. “We have automated this process by using drones to photograph the landscape and then examining the vegetation in the photographs,” explains Albrecht. “When the image recognition detects giant hogweed, we mark the place on a map.”

Neural network

Besides the drones, the project uses Artificial Intelligence, Machine Learning, and Big Data technologies. “The greatest challenge was to supply the algorithm with sufficient images to train the neural network to recognize the plant,” remembers Albrecht.

The project was conducted in collaboration with the Danish drone manufacturer ProDrone and involved four Danish town councils. The users update the existing mapping of the weed with the results of their IT-based image recognition and use the information to carry out control work.

Send HG Insights your news

Have you got IT news you would like the HG Insights news team to cover? Maybe you have the inside track on a major IT news story or have heard something significant or of interest to the global IT sector. We value all contributions.

Email all press releases / information to news@hginsights.com and one of our journalists will get back to you.

Back to News Index