With the increase of Artificial Intelligence (AI) in everyday life, developing AI-specific hardware based on brain-inspired computing is of utmost importance for efficient, adaptative and low-power systems. Neuro-inspired computing systems emulate the human brain's neuronal functions to efficiently solve problems that are easy to humans, such as pattern recognition. In this context, the EU H2020 NeurONN project explores a new energy-efficient computing paradigm based on phase-computing Oscillatory Neural Networks (ONN). It aims to create a neurocomputing chip that can be deployed on edge devices for AI.In this talk, a novel and alternative neuromorphic computing paradigm based on oscillating neural networks (ONN) will be presented. Energy efficient relaxation oscillators based on phase-change VO2 material for oscillating neurons and tunable 2D TMD MoS2 memristors for synapses are the building blocks of ONN architecture. Inspired by neural oscillations or brain waves, in ONN, the information is encoded in the phase of coupled oscillators. The talk will cover aspects from materials, devices, circuits to ONN architecture design and hardware implementation and demonstration on AI tasks. To demonstrate the ONN operation, we create a robotic application using two ONNs serially (ONN 1 feeds ONN 2), configured for pattern recognition to perform obstacle avoidance. We use a robot equipped in the front with eight infrared proximity sensors.