Research projects

Brief description: Development of an infrastructure for the reduction of emissions in intensive livestock farming and for the storage, processing and derivation of sensor values

Head: Marius Khan

Duration: 05.2016 - 06.2018

Funded by: Federal Ministry for Business Studies and Energy(Opens in a new tab) 

Funding by: ZF4038203SA5

The aim of the project is to develop an infrastructure for the permanent monitoring and subsequent reduction of harmful gas and dust pollution in intensive livestock farming stables. On the one hand, this is geared towards the requirements of Business Studies with regard to the most economical and therefore competitive stable operation possible. On the other hand, a significant reduction of up to 50% in ammonia pollution in the stable air is expected, which will greatly benefit human and animal health and at the same time reduce the need to administer medication. Furthermore, a high degree of purity of the exhaust air from stables increases the approval of the population with regard to the construction and expansion of animal stables in the vicinity of residential areas. Within the scope of the project, the initial focus is on the collection of harmful gases and ammonia reduction in pigsties. However, the approach is to be designed in such a way that it can also be transferred relatively flexibly to stables for other animals, such as cattle.

The use of the technology to be developed should, for the first time, enable a sustainable reduction in harmful gas pollution, which will also be documented and therefore verifiable. Permanent documentation of harmful gas pollution also results from the new legal basis within the European Union for the continuous monitoring of harmful gases in animal stables.

In summary, there are three main technical focuses:

Development of a sensor system for harmful gases, in particular ammonia, based on a new infrared measuring method (IR measuring method), which is suitable for special use in animal stables. The sensor should be able to be used both as a stand-alone gas measurement system and as an integral component of stationary or mobile measurement systems for the stable atmosphere.

Development of a sensor board which, in addition to the detection of ammonia, also enables the connection of other sensor types during operation. It should be possible to connect it to a hardware/software control system supplemented by suitable interfaces to control the components in a barn that influence the climate (filters, fogging system). The cooperation partner Barntec UG is responsible for the development of the sensor technology, the sensor board and the hardware/software control and is an experienced company partner in the field of sensor development in agriculture.

Development of a software-based smart data infrastructure that receives the measured sensor data of a barn from the hardware/software control system, derives control rules for an optimal barn climate based on statistical models and machine learning techniques and also allows long-term documentation of the harmful gas load. Based on the permanently transmitted information on measured harmful gas concentrations and additional climate-determining data, the smart data infrastructure continuously adapts the control rules depending on the situation and transmits them to the hardware/software control system. By incorporating existing infrastructure, the aim is to create an optimum climate in the barn by reducing harmful gas pollution as much as possible.

Figure 1 illustrates an overview of the system to be developed. IDiAL is responsible for developing the smart data infrastructure.

Brief description: High-performance sensor technology with cloud-based real-time data processing for the digital road in urban and long-distance traffic

Head: Fabian Wackermann

Duration: 08.2015 - 12.2017

Funded by: Federal Ministry for Business Studies and Energy(Opens in a new tab) 

Funded by: ZF4038201DB5

One of the major challenges facing society is the transformation of the transportation system. In addition to increasing demands on road users, such as the efficient utilization of traffic routes depending on the traffic situation by means of autonomous and intelligent vehicles, the transport infrastructure in particular must also be taken into account.

In the research project "High-performance sensor technology with cloud-based real-time data processing for the digital road in urban and long-distance traffic" funded by the Federal Ministry for Business Studies and Energy, Dortmund University of Applied Sciences and Arts is working together with the Technical University and Wilhelm Schröder GmbH on problems relating to the digital road. The project partners are focusing on the comprehensive collection of traffic data in real time. This results in various application scenarios, including the instant detection and reporting of wrong-way drivers, demand-oriented traffic flow control in urban areas and traffic and parking space balancing. The prototype being developed in the project integrates the necessary sensor technology into delineators at the roadside.
This makes the costly installation of induction loops in roads obsolete.

The project team responsible at the University of Applied Sciences is developing the smart data platform required for processing, evaluating and linking the sensor data. This is part of the process shown in Fig. 1, which the project partners are implementing and which comprises the following steps:

- Step 1: Raw data acquisition
Using techniques based on radiotomography, the sensors integrated into the delineators collect temporal data on passing vehicles.

- Step 2: Detection of vehicle
characteristics
Various pattern recognition methods are used to determine the lane, direction, speed and vehicle type (truck, car, etc.). Based on the data collected, the detection algorithm is successively improved.

- Step 3: Communication
Using radio or wired transmission paths, the delineators send detected vehicle characteristics to the smart data platform.

- Step 4: Interface provision
The platform offers the option of flexible provision of interfaces for incoming and outgoing data. The project team uses model-driven software development techniques for this purpose. The aim is to create domain-specific languages that enable the connection of new communication channels during operation. Requirements such as data format and communication protocol are taken into account so that new sensor types or adapted transmission formats can be integrated into the platform at runtime.

- Step 5: Data processing
In order to be able to evaluate data such as detected wrong-way drivers in real time, even with high traffic and data volumes, the project team implements prioritized processing queues. These allow, for example, the immediate dispatch of warning messages in the event of a wrong-way driver. Like the interfaces, the calculation methods for data aggregation can also be adapted at runtime. For example, methods for calculating the number of passing vehicles or the number of vehicles in a parking lot per time unit can be flexibly integrated into the platform and made available via declared interfaces.

Short description: Enabling of Results from AMALTHEA and others for Transfer into Application and building a Community

Head: Carsten Wolff(Opens in a new tab) 

Duration: 09.2014 - 08.2017

Funded by: Federal Ministry of Education and Research - 01IS14029K(Opens in a new tab) 

Fundedby: 01IS14029K

AMALTHEA4public deals with software development for embedded multi-core systems mainly in the automotive domain, but can also be applied in other embedded domains. The main focus is the efficient and effective support of model-based software development for embedded multi-core systems.

AMALTHEA4public aims to integrate the results of various publicly funded projects into the methodology and Eclipse-based tool platform developed by the previous AMALTHEA project. The aim is also to strengthen the community and promote the use of the platform in industry and research institutions in order to establish the tool platform as a de facto public standard across the board.

As a result of the previous AMALTHEA project, an Eclipse-based and publicly accessible tool chain infrastructure is available, which already contains some basic tools. In addition, AMALTHEA4public offers new and simple options (interfaces) for making extensive adaptations and extensions to the platform. Already planned features are test applications, verification and validation, safety concepts (according to ISO-26262 and ISO-61508), product line development and many-core supporting development processes, as well as domain extensions in the direction of ICT and automation technology.

Software projects in the field of embedded systems and above all in the automotive sector are carried out by automotive manufacturers (OEMs), suppliers, tool providers, software component developers or various engineering and consulting companies. The increasing complexity of these projects requires customized tool chains that are tailored to the needs of the respective project. The tool chains combine commercial, proprietary and open source modules. AMALTHEA4public provides a basic infrastructure for these tool chains in order to connect the tools and enable consistent data management. In particular, this also significantly promotes cross-company research and development as well as interdisciplinary collaboration, so that complex products can be developed more efficiently and tool providers can integrate their products more easily. This improves the reusability and joint use of development modules.

Innovation in AMALTHEA4public lies particularly in the integrative concepts of the open source tool chain platform and the definition and integration of all necessary tools in order to cope with the considerable scope and increasing complexity associated with multi-core ECUs. Interfaces, models or even DSLs from other publicly funded projects can be designed and developed for integration or examined and adapted. Innovative concepts can be integrated into AMALTHEA4public at early stages of development to enable early adaptation impact, behavioral analysis, verification and testing phases and similar techniques.

The most important results are the Eclipse-based tool chain platform, the integration of tools for all important development phases, as well as the demonstrators for presenting the possibilities in industry and research using the expandable AMALTHEA design flow. In addition, a community is to be created around the open source environment in order to disseminate results and ensure the continuous further development of the platform.

Brief description: Vital data at a glance

Head: Jan Oelker

Duration: 06.2016 - 05.2017

Funded by: Ministry of Innovation, Science and Research of the State of NRW(Opens in a new tab) 

Funding by: EFRE-0400075

The nursing crisis is a global problem, as there will be a shortage of 40 million healthcare professionals worldwide by 2030. The reason for this is the weak wage development and the increasing demand as a result of the demographic development. This trend will make senior citizens the largest demand group, which will put considerable pressure on the healthcare system.
This forecast is worrying for care services, relatives and those affected. Nevertheless, many people do not want to leave their home environment even if they need care and want more control, independence and active participation in the healthcare system. As a result, there is an increased demand for technical support. However, current support systems require the cooperation of the user,
are manufacturer-specific and do not take privacy into account.

covibo is a technical all-round carefree
carefree system that enables older people to live independently in their own homes for longer. It improves care, safety and comfort without interfering with everyday life. At the same time, the system relieves the burden on relatives and the care service. The focus is always on high benefit and ease of use.

covibo combines the three basic functions of automatic vital data recording, activity tracking and therapy plan.
With automatic vital data recording, the system accesses measured values, such as weight and blood pressure, directly after a measurement and saves and documents them. This means that no further action is required in addition to the usual measurement. Residents' activity is recorded passively via motion detectors, door contact sensors and the operation of the system. A call for help is sent out in the event of conspicuous inactivity.
All medication taken and measurements are entered in the therapy plan so that reminders sound and no action is forgotten. The sensitive data is stored locally at home. Residents can grant relatives, the care service or doctors access to certain data. The system consists of a compact base station, a mobile app and a web service. The base station includes all functions and interfaces for the surrounding devices and serves as a digital file folder. All data can be viewed visually via the app. If the resident refuses to use the app, they can use the system without having to operate it.
The web service, which is not visible to the end user, enables external access to the data.

The area of application is in private homes in collaboration with relatives or private domestic helpers. This approach is a response to the growing demand for flexible services in the home environment. The possible cooperation with the care service extends the scope of application. Which form of care is most suitable depends on
the need for care and the needs of the person concerned. Bluetooth Low Energy has successfully established itself on the market as a short-range wireless technology, which is why it is also used in covibo. The number of measuring devices and sensors for this technology is constantly increasing. The choice of devices to be integrated is manufacturer-independent, resulting in an open, easy-to-integrate system. covibo is currently being used and tested under real conditions together with a care service.

Brief description: Eyetracking-based interaction management of synchronous written communication

Head: Andrea Kienle

Duration: 04.2015 - 03.2017

Funded by: DFG - German Research Foundation(Opens in a new tab) 

Funding by: GZ KI 864/3-1

The DFG project eyetracking-based interaction management of synchronous written communication (ebiss) designs and tests eye tracking as the basis of innovative interaction management and attention arousal for synchronous written communication.

The starting point of the project is the observation that communication systems to support synchronous, written communication are established and widely used despite a variety of alternatives such as video conferencing systems. Problems often arise in the course of interaction, as participants can write at any time and especially in parallel. Conventions familiar from oral conversation and the many implicit signals that participants exchange with each other are only partially applied in synchronous written communication. In synchronous written communication, these cues are missing and a communication situation arises in which new rules for interaction must be established.

One way to establish the interaction management of oral conversation in synchronous written communication is to expand the technical systems with the aim of providing participants with missing information and offering more comprehensive interaction options.

Eye tracking, i.e. the recording, evaluation and reflection of the communication partners' eye movements, represents an additional channel for the transmission of human actions in synchronous written communication in order to support interaction management.

The evaluation and reflection of eye movement data incorporates communication patterns developed in linguistics. Communication patterns describe a structured sequence of communication processes. The aims of the project are therefore to analyze the special features of interaction management in synchronous, written communication and, based on this, to develop validated design recommendations for eye-tracking-based communication tools. This basic research is of great interest in view of the current widespread use of eye-tracking hardware.

In two laboratory studies, different display variants of the eye movement data were evaluated. It was shown that the subjects tended to refer to each other more in their contributions by reflecting back the eye-tracking data, resulting in greater convergence in communication.

Brief description: Development of a self-localization system for determining the exact position and orientation of mobile ground-based systems based on multi-sensor data fusion

Head: Christof Röhrig  

Duration:06.2014 - 11.2016

Funded by: Federal Ministry for Business Studies and Energy(Opens in a new tab) 

Funded by: KF2795209RR3

Automated guided vehicles (AGVs) and mobile robots are used for the automated transportation of goods and must be able to navigate automatically, i.e. without human intervention, in order to perform these tasks. Traditionally, AGVs are track-guided by means of an optical guide line or inductive guide wire. This method is very inflexible, which is why nowadays AGVs are primarily guided using virtual guidance. Mobile robots generally have a higher degree of autonomy than AGVs and navigate completely freely, i.e. without physical or virtual guidance. For navigation, AGVs and mobile robots require knowledge of their own position and orientation (orientation)
in relation to a two-dimensional coordina
system of the operational environment (self
localization / self-localization). The localization of mobile systems outside buildings can be achieved by using satellite positioning systems such as GPS. In environments in which no or only insufficient GPS information is available, e.g. inside buildings or in areas shadowed by buildings, it is not possible to determine the position using satellite positioning. In such environments, other positioning technologies must be used. One possible method for determining position using Auto-ID technology is grid navigation. This involves placing passive RFID transponders in or on the ground. Future-Shape GmbH has developed NaviFloor®, a floor covering in which RFID transponders are embedded. By embedding the RFID underlay in a synthetic resin floor, it is also suitable for high loads, such as those caused by heavy AGVs. Using an RFID reader, an AGV or a mobile robot can read the ID of an RFID transponder as soon as it is within reading range. By knowing the position of all RFID transponders in the operating environment, the AGV or mobile robot can determine its own position. Direct determination of orientation is not possible with this technology. Only when at least two RFID transponders have been passed over can the orientation be roughly determined.
To navigate AGVs using RFID transponders, the starting orientation has so far had to be determined manually and transferred to the AGV control system. Adjusting the orientation during the journey using RFID transponders requires additional sensors (wheel sensors, gyroscope) as well as precise knowledge of the kinematics of the AGV in combination with the geometry and attachment of the RFID antennas. Adaptation must be carried out individually for each AGV type and implemented in the vehicle control system. This adaptation effort is currently preventing a wider distribution of the NaviFloor®.
The aim of the project is to develop a modular localization device that will make it possible to determine the exact position and orientation of mobile ground-based systems such as automated guided vehicles (AGVs) or mobile robots outdoors using GPS and indoors based on the NaviFloor®. The device is designed to work independently of the sensors of the AGV or mobile robot. The core idea is to separate the hardware for localizing the AGV from the hardware for
the hardware for controlling the AGV in order to minimize the integration effort for AGV manufacturers. Together with Future-Shape GmbH, a device is being developed which, in addition to an integrated GPS and an RFID reader, will have further sensors for determining orientation and for precise position interpolation. By appropriately fusing the sensor readings with each other and with the data from the RFID reader, it should be possible to accurately determine the orientation and position.
The localization system is primarily intended for use in AGVs or mobile robots. However, all other mobile objects where a reader can be attached close to the ground are also supported. These can be, for example, forklift trucks, shopping carts, electric wheelchairs, hospital beds or even valuable devices or objects whose location is to be detected automatically.

Brief description: Development of a swarm-based localization system

Head: Christof Röhrig  

Duration: 06.2014 - 11.2016

Funded by: Federal Ministry for Business Studies and Energy(Opens in a new tab) 

Funding reference: KF2795208ED3

The localization of objects outside buildings can be realized by using satellite positioning systems such as GPS. In environments in which no or only insufficient GPS information is available, e.g. inside buildings or in areas shadowed by buildings, it is not possible to determine the position using satellite positioning. In such environments, other positioning technologies must be used. One possibility is localization via radio. Up to now
the prerequisite for radio localization is the establishment of a localization infrastructure.
Localization is carried out, for example, by means of runtime measurements between the fixed radio stations (reference nodes) and the mobile nodes to be localized. The measured transit times to at least three fixed nodes can then be used, e.g.
the position of a mobile node can then be determined, e.g. by means of trilateration.

In many applications, however, it is not possible to fall back on an infrastructure with reference nodes whose positions are known or to install them permanently. For example, the fire department has no way of setting up the necessary infrastructure before an operation in the event of a fire or disaster. If the area of operation is inside buildings, satellite positioning is also not available. The installation of a radio infrastructure is also very cost-intensive.

Together with its partner Nanotron Technologies, the swarmLOC project is developing a new type of infrastructure-free or low-infrastructure localization system that enables the cooperative positioning of people or objects using radio time-of-flight measurements. The system is to be based on a mobile ad-hoc network and therefore does not require any infrastructure to be installed. This approach makes it possible for such a system to be operational in a very short time.

Each radio node in the network can be optionally expanded with additional sensors. For example, GPS or proximity-based sensors, such as RFID, provide absolute position information,
which can supplement the relative positions of the cooperative localization system. Inertial sensors can also contribute to an increase in localization accuracy. Cooperative localization algorithms based on probabilistic filters are to be used to fuse the various types of sensor information. In particular, swarm-based algorithms are to be analyzed with regard to their performance and applicability in the planned system.

Figure 1 shows collision protection in opencast mining as a possible application of swarmLOC. The person carries a mobile radio node and communicates with the radio
radio nodes in the vehicles. Through cooperative localization, all radio nodes are localized relative to each other and the excavator drivers are warned if there is a risk of collision.

Brief description:
Flexible, two-stage Organic Rankine Cycle Turbine for the demand-driven generation of electricity, heat and cooling from waste heat

Head: Klaus-Peter Priebe

Duration: 08.2014 - 07.2016

Funded by: Federal Ministry for Business Studies and Energy(Opens in a new tab) 

Funding code:
KF3338401ST4 Fachhochschule Dortmund
KF2795210ST4 Fachhochschule Dortmund

The need to make the best possible use of waste heat potential for downstream electricity, cooling and useful heat generation is undisputed. On the one hand, global warming with the need to reduce CO2 emissions and, on the other hand, the foreseeable end of fossil fuels with foreseeable shortages and price increases should be mentioned here. In the ZIM-ORC project, a two-stage ORC (Organic Rankine Cycle) plant for the conversion of unused waste heat into electricity is being developed with funding from the German Federal Ministry for Business Studies and Energy (BMWi), which is technically particularly efficient and therefore economically viable. Organic Rankine Cycle turbines can efficiently generate mechanical work or electricity from low-temperature heat. As part of the project, a fully functional small power plant is being built as a prototype. Fachhochschule Dortmund University of Applied Sciences and Arts is developing the electrical and communication nervous system and the brain of the plant for optimal power and heat demand-driven control. The University of Paderborn is also involved in the project for the thermodynamic design, Smart Mechatronics GmbH for the control technology and Lütkemüller GmbH and Heim Präzisionstechnik GmbH for the mechanics of the plant and the turbines.

The design of the small power plant follows a concept of interchangeable modules - in Figure 1, from right to left, Module 1 with the direct evaporator is shown first, followed by Module 2 with the high-temperature working circuit with working temperatures of up to 300 °C,
module 3 with the low-temperature circuit and working temperatures up to 110 °C and finally module 4 with the direct condenser. Thanks to the modular design, systems can be built with one or two ORC turbines so that they can be adapted to different waste heat profiles. In addition, the system can operate with waste heat from an output of 500 kWth, whereas previous solutions only become economical with much higher heat outputs.
The development of the demand-driven control of the two-stage system represents a particular challenge. For this purpose, a hierarchical control and regulation system based on the Operator Controller Module (OCM, from the SFB614 at Paderborn University) is being developed. Here, the control in the first level is based on the thermodynamic data of the project partner University of Paderborn for the individual material cycles and the thermodynamic behavior of the individual devices. The control in the second level of the motor control loops was realized model-based using Matlab/Simulink in the form of a Model Predictive Controller (MPC). On level 3 of the OCM model, the communicative nervous system is implemented in the form of the reflective operator with the sequence controls and fault handling, e.g. to automatically intervene to correct faults or shut down the system in a controlled manner. Finally, the fourth level of the cognitive operator is the self-learning optimization program for the yield-optimized operation of the system with changing requirements, including maintenance cycles and costs.

The target markets for this technology can be found wherever waste heat is present from 300 °C with a waste heat output of more than 300 kWth. These heat sources are found in biogas plants and in industry as well as in solar thermal energy generation.
The development goal is to generate the additional electricity from waste heat to the condition of "better than grid parity",
i.e. 1 kWh from waste heat must not cost more than € 0.12 per kWh in the full cost calculation without subsidies (with a payback period < 6 years).