Energy management made in Darmstadt

No matter whether it’s in industrial halls, residential homes or public buildings, the supply of useful energy such as heating, cooling or compressed air is always reliant on the interplay between a variety of different technical systems. And despite the fact that these technical systems are becoming increasingly complex due to the switch to renewable energies and greater requirements for efficiency and sustainability, conventional control strategies are still overwhelmingly used for these systems. “These strategies are basic, rule-based and tried-and-tested control concepts but they are far from optimal”, says Dr.-Ing. Niklas Panten, Managing Director of etalytics. The processes employed up to now often fail to take into account time-based and stochastic variables such as the outdoor temperature, return flow temperature or relative humidity although these factors can have a huge influence on efficiency – such as in a cooling tower. They also fail to take account of efficiency losses that occur in the energy supply cascade to the consumer every time energy is converted.

A toolbox of possibilities

In order to make the energy supply more efficient and controllable in near real-time, Panten and his co-founders Thomas Weber and Björn Scheurich utilise the data that is produced on and ongoing basis in the technical systems. Numerous sensors generate a huge amount of data that provides information on temperatures, pressures, volume flows and other parameters that can be used to optimise the system. A small industrial computer taps into this data and streams it to the cloud where any gaps are removed and the data is applied and evaluated. Finally, the collected data is used to calculate optimal control signals for the actuators in the energy system. The optimised control data is sent back to the system directly from the cloud and is also used for, amongst other things, the visual representation of recommended actions. Manual interventions into the control software are thus no longer required. It is a novel, flexible and highly automated process that is designed to support energy managers in their work.

This process would not work without the AI solutions that Panten and Weber worked on for many years in the ETA research factory. “We have an entire toolbox of possibilities at our disposal”, says Weber. The company not only uses mathematical optimisations and processes based on monitored learning but also utilises learning systems that can make predictions about the future behaviour of an energy system.

From the idea to the start-up

Niklas Panten experienced the urgent need for innovation in the area of energy management for himself when he worked together with his research group on the development of the ETA factory at the Lichtwiese Campus from 2014 to 2019. The software solutions for data-based energy management that were available on the market at the time all proved too complicated and inflexible. Panten was convinced that “there had to a simpler solution with more functionalities”. At the beginning of 2019, he decided together with Thomas Weber to not only develop a better, AI-based software but also to sell it on the market. “Naturally, we were not able to achieve this goal with our data science skills alone”, explains Weber. And this is when the software architect Björn Scheurich joined the team.

They made an appointment for an initial consultation at the HIGHEST Startup and Innovation Center, learned about the range of services available to them and ultimately submitted an application for an EXIST Business Start-up Grant with support from the HIGHEST consultancy team. This grant is still covering the living costs for the entrepreneurs, work contracts for external developers and the rooms used by etalytics in the Energy Center at TU Darmstadt until the end of March 2021. Although work on the business plan and technology moved at a rapid pace right from the very beginning: “We often asked ourselves whether it was all worth the risk”, remembers Panten. The decisive motivational boost required to found the company was ultimately the team’s victory in the “Digital Innovations” start-up competition organised by the Federal Ministry for Economic Affairs in autumn 2019.

Next step: the first industrial installation

The etalytics team – which now has 14 members – will still have to work a lot of nights and weekends to achieve its goal of launching itself on the market with the etaONE® platform, which contains freely configurable modules for data collection and modelling, system optimisation, simulation and forecasting. The business model behind the platform: etalytics aims to host the platform as a Software as a Service (SaaS) solution on servers in the EU for its increasingly cloud-oriented customers, while an on-premise solution will also be offered to those customers who want to keep their data in-house.

etalytics GmbH aims to generate income from licences and the leasing of its etaEDGE® IoT gateway. A computer centre in Frankfurt is already onboard as a pilot user and a large automotive manufacturer has also expressed an interest. If the first industrial installation is placed into operation as planned in summer this year, etalytics will have achieved another important step on its path to growth.