Projects
1. ONGOING PROJECTS
High per Asphalt - High-performance asphalt with composite-optimized textile reinforcement and integrated sensors
The R&D collaborative project of the Saxon SMEs Dresden Elektronik Ingenieurtechnik GmbH, TECHiFAB GmbH, Wolf Straßen- und Tiefbau GmbH, Johne & Groß GmbH, as well as the research institutions of TU Dresden, the Institute of Textile Machinery and High-Performance Textile Materials (ITM), and the Institute of Urban Construction and Road Engineering (ISS), aims to explore a globally unprecedented, durable, sustainable, and at the same time intelligent, textile-reinforced asphalt construction method with integrated fiber-based sensors. Furthermore, resource-efficient in-situ data evaluation and analysis for AI-supported digital load monitoring is to be enabled. The goal is the experimental demonstration of functionality as well as validation in the laboratory of the novel technological approach.
TECHiFAB takes on the design and development of a chip for a system for resource-efficient sensor data processing for continuous and intelligent monitoring of road loads. A central goal is the design and development of novel data pre-processing algorithms that provide precise information through signal processing techniques and enable energy-efficient sensor data compression (directly on-site) for IoT-based applications.
AWEO - Atomic Layer Deposition (ALD) of bismuth- and iron-containing oxide layers
TECHiFAB GmbH and the Technical University of Dresden (TUD) are working together in a close research cooperation, funded by the Saxon Development Bank, on the development and optimization of industry-compatible manufacturing processes for bismuth iron oxide layers for innovative electronic applications.
The research and development project includes process development and optimization, material characterization and quality control, and the industrial applicability of producing BFO layers as a material basis for TECHiFAB GmbH’s novel, reconfigurable analog components.
The grant was provided from funds of the European Regional Development Fund (ERDF). The project was financially supported by the European Union and the Free State of Saxony.
lnnASaaS – Innovation Assistant for Software as a Service
In times of just-in-time production and tightly scheduled infrastructure utilization, machine and infrastructure failures can cause serious economic and health-related damages. Predictive maintenance uses artificial intelligence, specifically machine learning algorithms, to detect signs of impending failures early. These systems require real-time data from external or internal sensors. Standard practice involves an algorithm trained in neural networks to compare actual and target values. However, the decision-making processes of neural networks often function as a black box, as their logical steps remain unknown.
TECHiFAB has developed the COR-RiSTOR data processing algorithm to compare actual and target values while maintaining transparency in the logical decision-making process. In March 2024, the COR-RiSTOR demonstrator was introduced to the market. This project aims to strengthen TECHiFAB’s product-line-based Software-as-a-Service (SaaS) business and provide customers with an easy and low-effort digital access to COR-RiSTOR technology.
The funding is provided by the European Social Fund Plus (ESF Plus). The project is financially supported by the European Union and the Free State of Saxony.
2. COMPLETED PROJECTS
CLA-RiSTOR – Real-time Classification of Sensor Data with TiF Memristors
Based on central hardware components – the CLA-RiSTOR board and the TiF-MEMRiSTOR chip with 32 TiF-MEMRiSTORs – an initial small-scale series of the CLA-RiSTOR Classification Kit will be produced and introduced to the market. The CLA-RiSTOR Classification Kit will demonstrate the principle of analog, brain-inspired computing for classification tasks to manufacturers of electronic hardware modules for data classification. Relevant application fields include autonomous driving, neuro-signal processing, medical image recognition, and data quality monitoring.
This initiative was co-financed with tax funds based on the budget approved by the Saxon State Parliament. The funding was provided by the European Regional Development Fund (ERDF). The project was financially supported by the European Union and the Free State of Saxony.
Business Angel Bonus – Market Expansion for Electronic Products with Highly Innovative TiF Memristors
TECHiFAB aims to expand the market for electronic products featuring highly innovative TiF-MEMRISTORs, which are already marketed as discrete individual components. Several electronic product lines incorporating TiF-MEMRISTORs have been conceptualized, including COR-RiSTOR, ALU-RiSTOR, PUF-RiSTOR, and CLA-RiSTOR. Following the market launch of COR-RiSTOR, ALU-RiSTOR is being developed for neuromorphic chips and neuromorphic computing. To achieve these goals, the company will be further developed as part of the project.
The project is financially supported by the European Union and the Free State of Saxony.
Development of an Online Shop and Integration with Existing ERP System
The newly developed digital platform offers a time- and resource-efficient way to sell TECHiFAB products via an online shop integrated into the TECHiFAB website. It features automated order and payment processing, customer data management, invoicing, and controlling functions through the TECHiFAB ERP system. Additionally, the system enables access to a website-based Software-as-a-Service offering, directly linked to the available TECHiFAB hardware in the form of memristor boards.
The funding was provided by the European Regional Development Fund (ERDF). The project was financially supported by the European Union and the Free State of Saxony.
COR-RiSTOR – Real-Time Correlation Analysis with BFO Memristors
The project aimed to produce an initial small-scale series of the COR-RiSTOR Correlation Kit, establish appropriate marketing strategies, and professionally launch the product. The COR-RiSTOR Correlation Kit is the world’s first analog AI hardware for real-time correlation analysis (<1 ms), based on genuine memristor technology. It is designed for both linear and nonlinear correlation analyses of edge sensor data and achieves up to 100% recognition accuracy.
Promising applications include data evaluation, future prediction, predictive maintenance, autonomous systems, smart homes, and intelligent assistance applications. The target audience was addressed in three phases. The market launch was accompanied by a Kickstarter campaign to promote the product worldwide via digital media.
This initiative was co-financed with tax funds based on the budget approved by the Saxon State Parliament. The funding was provided by the European Regional Development Fund (ERDF). The project was financially supported by the European Union and the Free State of Saxony.
Investments in TECHiFAB GmbH Infrastructure
At the time of the project, industrial-scale production of TiF-MEMRISTORs capable of simultaneously processing and storing analog data was not feasible. Investments were made in TECHiFAB GmbH’s infrastructure to develop manufacturing processes for these memristors, with the goal of enabling small- and medium-scale production.
This initiative was co-financed with tax funds based on the budget approved by the Saxon State Parliament.
ALU-RiSTOR
The project aimed to develop an optimal design for a discrete circuit with TiF-MEMRISTORs and a software interface for performing fundamental arithmetic operations with multibit and real numbers without data transfer. In the long term, ALU-RiSTOR technology serves as a building block for resource-efficient neuromorphic computing architectures.
The funding was provided by the European Regional Development Fund (ERDF). The project was financially supported by the European Union and the Free State of Saxony.
Development of the Circuit Design for a Memristor Test Kit
The Memristor Test Kit was developed to characterize and demonstrate the properties of TiF-MEMRISTORs. Testers can use the kit’s software interface to apply voltage to the memristors, change their state (write voltage), and read their state (read voltage). The memristor’s state is typically analyzed based on its resistance.
The funding was provided by the European Regional Development Fund (ERDF). The project was financially supported by the European Union and the Free State of Saxony.