Neuromorphic Computing and its underlying theoretical foundation: spiking neural networks (SNN), have been a research topic for decades [1]. Nevertheless, the industry didn’t pick it up massively yet, because of the already steep ramping up they face with conventional AI, because dedicated hardware reached commercial viability recently, and also because neuromorphic evangelists aren’t that many. We, at fortiss, have the mission to transfer neuromorphic research to the industry, and prove that neuromorphic computing makes a point by fixing real industrial problems. We will show that a complete neuromorphic workflow can be assembled to solve the complex industrial task of cable insertion with a robotic arm [2]. The integration of a full neuromorphic task pipeline is unprecedented, due to the complexity of handling task transitions in a pure SNN implementation. We will show that energy-efficient and low-latency gesture recognition with event-based (aka. neuromorphic) cameras enable early and adaptive inferences for applications in sports[3], arts[4] and defence. Applications in object tracking for logistics, anomaly detection in satellite communications and welding cobots will also be showcased. These examples foster various hardware and event-based [5] sensors, as well as different architectural approaches. They use the advantages of SNN extensively, with on-chip online learning. They prove that the neuromorphic ecosystem is ready for industrial applications and that a use-case driven approach can lead to technological progress too.