CERES aims at evaluating possible commercial applications of new protein-based down-converting red-emitting LED sources. Laboratory prototypes exhibit an extraordinary stability and efficiency as well as an easy-to-tune emission spectrum. This contrasts with the commercial low-energy emitting LED technology. CERES will focus on a validation phase including i) the optimization of the upscaling of protein production, the preparation of large-area protein-polymer color filters, and the assembly of red-emitting LED arrays and ii) the test in pre-industrial crop controlled environment agriculture growing boxes. This information will be paramount to realize a realistic market impact.

This proof-of-concept project aims to develop and bring to market a novel visuo-haptic sensor system for robotic manipulators, which allows for multimodal sensing, reduced system complexity and significantly lower costs compared to current systems. It replaces dedicated force sensors with passive components and a camera, providing coherent measurements of both force and contact shape. Three sensor setups for grippers, mobile platforms and tools on the endeffector have already been implemented as research prototypes in the context of the ERC Starting Grant ProHaptics. The commercial prototype to be developed within this project will consist of a standard commercial gripper, the visuo-haptic sensor, as well as soft-ware for manipulation planning. In later stages, it is planned to apply the concept to an entire robotic arm, replacing the dedicated and expensive joint and torque sensors required today. Costs and system complexity are cut considerably by our approach since rigid mechanics as well as highly specialized sensor systems as used by current robot arms are no longer required. The application focus of this proof-of-concept project is collaborative production. This concept helps to keep production competitive in in high-income countries. While our market entry strategy targets the commercially highly relevant production scenario, we believe that this paradigm will open new markets for robotic systems also in interpersonal communication, household robotics and games/entertainment. Such systems need low-cost multimodal sensor systems, which provide a rich representation of the environment. In a recent market study, a high interest for robotic manipulators that operate in un-structured environments and offer natural compliance has been identified, especially in the application areas of robotic commissioning and joint mounting, verification and documentation of components.

With the recent breakthrough of deep learning methods, we currenty see the advent of employing this methodology in the context of physical simulations. Such simulations are widely used in numerous industrial fields, starting from car and airplane manufacturers, over computer graphics and animations to medical blood flow simulations. The market for computer simulations is currently exceeding 15 billion USD world wide, with rising trends, and 3 billion spent in Europe alone. A significant fraction of these simulations focuses purely on solving various forms of the Navier-Stokes equations. While right now virtually all of these simulations use traditional solvers, we estimate than only a few years from now there will be a significant fraction of deep learning powered solvers. Thus, we are at the right point in time to lay the foundations for commercializing the technology of deep learning for fluid simulations. The goal of this PoC project is to develop a first commercial flow solver based on deep learning that can predict fluid flow solutions almost instantly using a pre-trained model. This project will enable the team of Prof. Thuerey to mature the algorithms developed as part of the ERC Starting Grant \realflow, and turn them into the basis of a marketable product. The initial models will be thoroughly tested and validated, in order to satisfy industrial requirements for reliability and accuracy. In addition, this PoC aims for establishing a platform for flow data collection, interface standards, and trained models. This platform will be developed in conjunction to the deep-learning powered flow solving application, and provide research connections and publicity in parallel to it.