WIN-LAB started the Big Power Data project included in the 7th Energy Program of FFG. HAKOM TSM launched a cloud-first strategy to meet new requirements in the energy field resulting from the challenges they face from the increasing share of green electricity in the European Power Grid. The main focus of the platform is on Time Series Data, but it fits KPIs as well. HAKOM TSM supports SQL, NoSQL, inMemory and in the near future NewSQL databases in the cloud and on premise. The main challenge is to scale out with ACID transactions. Therefore, some NoSQL solutions known for horizontal scalability based on BASE do not fit. That is where LeanXcale comes into play. LeanXcale is one of a few that supports linear scalability with fast ingestions and ACID transactions. Thus this strategy for the rapidly changin field of green energy. For more information click this YouTube Link

WIN-LAB On 7 January 2021 WIN-LAB started the data cleansing collaboration with the club plattform vereinsplaner.at to take it to the next level. verbandsplaner.com should support the clubs and as a new feature associations even better with their tasks. For that the offers for the clubs which can be modularly incorporated into their website or fully hosted on verbandsplaner.at will be expanded with the association feature. For the enormous data transfer verbandsplaner.com, the research lab AIST and the WIN-LAB in Hagenberg are developing together a morphable data importer which automates the data transfer as far as possible. To do this, AI models are trained and used which enable the data importer to automatically and constantly process new data. During the data transfer the data should also be cleaned and if necessary, enriched and brought up to the standard for the hosting at verbandsplaner.at.

WIN-LABWIN-LAB began to develop a community platform together with Afreshed GmbH as a part of a FFG Basisprogramm. This platform connects consumers, producers and logistics companies in order to save consumable food that otherwise would be burned in a biogas plant or composted. Yearly 800,000 tons of groceries that don’t meet that specific standards of the wholesale trade are sorted out. They launched a platform that simultaneously saves those groceries, allows farmers to get a better offer for their products and presents customers fresh food. This idea convinced the Jurors of the Austrian PULS4 show “2 Minuten 2 Millionen” to make an investment. The FHOOE also supports to stop this food waste.

WIN-LAB It is based on machine learning methods from the field of “Natural Language Processing”, which the research group has already successfully used in several projects, and which come from an area in which great progress has been made in recent years. The cooperation with the company is very close and fruitful, among other things because many graduates of our “Mobility & Energy” department are with the company.

WIN-LAB New Deep Learning Hardware bought by the WIN-LAB Research Group, together with the Department of Digital Media, the Department of Mobility & Energy and the Department of Medical and Bioinformatics, is used for the creation of complex machine learning models. With the Deep Learning Server, we can now train our neural networks in Tensorflow, Torch and other deep learning frameworks in much less time. The Deep Learning Box is highly optimized for machine learning and deep learning applications. Equipped with eight of the latest NVIDIA RTX 3090 GPUs, it provides enormous computing power. Current areas of application are the creation of new language models, anomaly detection for risk assessment for companies and topic modeling of large text volumes.