Nicolas Schmid Email and Phone Number
Nicolas Schmid is a passionate researcher with a broad interest in machine learning, control theory, and system identification and their applications in various fields, including energy systems, robotics, spectroscopy, and neurosciences. His research involves combining novel theoretical and algorithmic concepts to develop practical solutions for real-world applications. Nicolas's interdisciplinary background and versatile interests enable him to combine knowledge from various fields, making him particularly interested in interdisciplinary research.Nicolas holds a Master's degree in Mechanical Engineering from ETH Zurich, specialising in artificial intelligence, control, and robotics. During his master's thesis, he applied state-of-the-art machine learning and control algorithms to enhance the safety and efficiency of real, safety-critical energy systems.Currently, Nicolas is pursuing a Ph.D. in Data Science, where he focuses on the development of machine learning methods for nuclear magnetic resonance (NMR) spectroscopy. His research project aims to predict the spin system of small molecules from high-resolution liquid nuclear magnetic resonance spectra using data-driven optimization and machine learning techniques while respecting NMR spectroscopy's physical and chemical structure. Through his Ph.D. studies, Nicolas is working closely with the Data Science team of Bruker Biospin AG, a project partner in an Innosuisse project, to automate the tasks of synthesis control and structure elucidation of NMR spectra.Nicolas is a member of a collaborative Ph.D. program in Data Science between the University of Zurich (UZH) and the Zurich University of Applied Sciences (ZHAW). He is supervised by a team of experts, including Prof. Roland Sigel, an NMR specialist at UZH, Prof. Jan Dirk Wegner, an expert in Machine Learning at UZH, and Prof. Dirk Wilhelm, who specializes in NMR and Machine Learning at ZHAW. This collaborative program allows Nicolas to integrate knowledge from different fields and contribute to solving complex problems that require interdisciplinary collaboration.His research aims to contribute to the area of NMR spectroscopy, making it more accessible and efficient for a wide range of applications by facilitating workflows for non-experts and assisting experts with challenging decisions.
-
Phd Student, Data ScienceUniversität Zürich | University Of Zurich Apr 2021 - PresentZürich, Schweiz -
Phd Student, Data ScienceZhaw Zurich University Of Applied Sciences Apr 2021 - PresentZurich, SwitzerlandMy PhD project focuses on predicting the spin system of small molecules from high-resolution liquid nuclear magnetic resonance (NMR) spectra using data-driven optimization and machine learning (ML) techniques. During my PhD, which is partly in the scope of an Innosuisse Project with Bruker Biospin AG, I closely collaborate with their Data Science team. We work on automating the tasks of synthesis control and structure elucidation of NMR spectra. I am enrolled in a collaborative PhD program in Data Science between the University of Zurich (UZH) and the Zurich University of Applied Sciences (ZHAW) where my supervisors are Prof. Roland Sigel (NMR, UZH), Prof. Jan Dirk Wegner (ML, UZH) and Prof. Dirk Wilhelm (NMR & ML, ZHAW). -
Research AssistantZhaw Zurich University Of Applied Sciences Oct 2019 - Apr 2021Winterthur Area, SwitzerlandDecision making and control, e.g. reinforcement learning, Bayesian optimization, data driven control etc., in complex systems with emphasis on the optimization of renewable energy systems and energy savings.
-
Phd Student, Data ScienceBruker Biospin Apr 2021 - Jul 2023Fällanden, Zürich, SchweizInnosuisse Project -
Civil ServiceEmpa Feb 2019 - Jun 2019Zürich Area, Switzerland -
Master Thesis: "Data Driven Adaptive Controller Parametrization: A Bayesian Optimization Approach"Eth Zürich Apr 2018 - Nov 2018Automatic Control Laboratory (D-Itet)In the most of industrial and commercial plants, low order controllers such as P-, PI- and PID-controllers are the current common choices. Since these controllers may not be properly tuned, the corresponding close-loop systems are not operating efficiently in terms of energy consumption, user’s comforts and monetary costs. This fact highlights the importance of tuning these controllers, preferably in an automated and model-free way so that they can be easily tuned specifically when the access and modification to the plant is very limited and minimal. Recently in the literature and in practice, data-driven approaches based on black-box and Bayesian optimization are utilized for toy examples such as inverse pendulum and applications like robotics and combustion engines. Following the same lines, we investigate tuning a controller implemented for a heat pump with unsatisfactory performance. We utilized an approach based on safe Bayesian optimization with Gaussian processes which was well in agreement with the given operational constraints and available resources. By applying problem-relevant performance metrics and safety constraints for the Bayesian optimization, we show that the overshoot and the mean squared tracking error can be reduced on average by more than 70% and more than 80%, respectively. For further investigations, the heat pump dynamics were emulated authentically and accurately with a data driven model. The closed-loop model of the heat pump consists of a recurrent neural network, disturbance model, PID controller and contextual-safe Bayesian optimization block. This way, we can go beyond region of safe parameters and systemically compare different hyper-parameter, Gaussian process kernels and settings of the contextual-safe Bayesian optimization. The generality of the data driven model might even allow us to use it for other heat pumps or similar systems with minimal additional effort. -
Vehicle Autonomy Engineer In TrainingDuckietown Engineering Co. Sep 2017 - Mar 2018Zürich Und Umgebung, Schweiz -
Teaching Assistant In Stochastics (Probability And Statistics)Eth, Swiss Federal Institute Of Technology, Zurich Sep 2017 - Jan 2018
-
Research AssistantEth, Swiss Federal Institute Of Technology, Zurich Jul 2016 - Sep 2017Eth Zurich, SwitzerlandSoftware Developement, Measurement Automation and Data PostprocessingResearch Assistant at Micro -und Nanosystem Group: Developement of Carbon Nanotube FET-based Gas SensorsCarbon nanotube chemical sensors promise incredibly low operating power compared to existing technologies, in addition to their small size. By analyzing the change the in electrical transport characteristics of single-walled carbon nanotubes (SWNTs) upon exposure to gaseous analytes (e.g. NO2), carbon nanotube-based field-effect transistor devices are employed as chemical sensors.Workload: 40%
-
Semester Thesis: Deep Reinforcement Learning For Battery Scheduling In Energy Prosumer SystemsEth Zürich Aug 2017 - Dec 2017Bits To Energy LabModern energy management systems urge for intelligence in terms of the energy handling and usage. We investigated the interplay between a real household with a battery and photovoltaics and the grid. The idea was to take intelligent decisions with respect to charging and discharging the battery in order to fulfill a certain purpose like peak shaving or profit maximization for the prosumer.Therefore we controlled the actions with respect to the battery with Reinforcement Learning. Incontrast to most common control algorithms Reinforcement Learning doesn't necessarily need amodel of the environment and is able to learn the optimal behavior implicitly. In the recent past,especially Deep Reinforcement Learning produced very promising results in complex environments, similar to ours.In order to use Reinforcement Learning with our environment, we had to model the deterministicstate dynamics of our system according to the actions taken with respect to the battery. Afterwards we created reward functions for the different purposes i.e. peak shaving and profit maximization. In the end we ran different RL algorithms, i.e the so called Q-Learnig algorithm and the state of the art Proximal Policy Optimization algorithm, on our environment with the according reward function. Especially in the case of peak shaving, we were able achieve good results. We reduce all peaks below a certain threshold with the Proximal Policy Optimization algorithm and a feasible reward function. This peak shaving behavior could help to omit the fossil energy intensive and costly peaks in the electricity consumption and possibly lead to better electricity price conditions for prosumers. Furthermore households which stress the grid much less than others by loading the battery with photovoltaic power may get a reduction on the grid usage fee, which builds a part of the electricity price.
-
Industry InternshipQuo Ag Mar 2016 - Aug 2016OerlikonProductdevelopement and technical consulting:-functional models -building and testing prototypes-feasibility studies-market analysis-brainstormig and innovation workshops-technical concepts and design -
Engineering InternshipGeorg Fischer Jan 2011 - Feb 2011
Nicolas Schmid Education Details
-
Mechanical Engineering
-
Mechatronics, Robotics, And Automation Engineering
Frequently Asked Questions about Nicolas Schmid
What company does Nicolas Schmid work for?
Nicolas Schmid works for Universität Zürich | University Of Zurich
What is Nicolas Schmid's role at the current company?
Nicolas Schmid's current role is PhD Student, Data Science.
What schools did Nicolas Schmid attend?
Nicolas Schmid attended Eth Zürich, Eth Zürich.
Not the Nicolas Schmid you were looking for?
-
-
Nicolas Schmid
Senior Project Manager At Infras | Climate & Energy Policy | Phd Eth ZurichSwitzerland -
Nicolas Schmid
Basel -
1roche.com
Free Chrome Extension
Find emails, phones & company data instantly
Aero Online
Your AI prospecting assistant
Select data to include:
0 records × $0.02 per record
Download 750 million emails and 100 million phone numbers
Access emails and phone numbers of over 750 million business users. Instantly download verified profiles using 20+ filters, including location, job title, company, function, and industry.
Start your free trial