Canton of Zurich's Innovation Sandbox for AI provides practical insights from five AI projects

(CONNECT) The Innovation Sandbox for AI of the Department for Economic Affairs of the Canton of Zurich has tested five AI use cases under real conditions, namely in the context of building permit procedures and medical documentation as well as inspection robots, sensor-based monitoring of bridges and the evaluation of retinal images. According to a statement, a total of 24 organizations applied to take part.

In the AI for building permits project, the Innovation Sandbox, together with project partners Gossweiler Ingenieure, Byte Studio and Nokema and in close coordination with the building department, has developed a prototype for an AI-supported preliminary check in the notification process, for example for solar systems, heat pumps or charging stations. This helps applicants to recognize at an early stage whether the documents are complete and where formal details are missing or need to be improved. "The project on building permits is an example of how the responsible use of AI can contribute to better submissions and greater efficiency - without weakening the decision-making authority and responsibility of the authorities," said Carmen Walker Späh, Government Councillor and Head of the Department for Economic Affairs.

An example from the company MPAssist was used to investigate how AI can relieve the burden on doctors when transcribing and structuring medical reports. Together with the Eye Clinic of Zurich Triemli City Hospital and the Werner H. Spross Foundation, it has been investigated how AI can support the evaluation of retinal images in clinical practice. The results of the project with irmos technologies and Schweizerische Südostbahn clearly show that sensor data is a useful addition to traditional on-site inspections of railroad bridges. Together with ANYbotics, the manufacturer of four-legged inspection robots, the regulatory conditions under which robotic systems can be used for inspections in critical infrastructure such as energy plants were analyzed.

"The second phase of the Innovation Sandbox for AI shows that artificial intelligence creates added value above all where it supports specific, practical tasks," says Raphael von Thiessen, Program Lead AI Hub at the Office for Economy. "Widespread use therefore requires not only powerful models, but also test environments that combine the technology with clean data, clear processes, human control and a reliable legal framework." ce/mm