Sensor technology research area

Development of a system for radar-based detection of wild animals at the roadside (OhDeer)

Project manager: Prof. Dr.-Ing. Thomas Walter and Prof. Dr. Hubert Mantz
Project duration: 01.09.2022 - 30.08.2024
Funding: Federal Government - BMVI
Program name: mFUND - Data Innovations for Mobility 4.0
Project description:
The aim of this project is to develop an intelligent wildlife detection system. The radar-based detection of wild animals enables increased protection of vulnerable road users. In the project, THU is developing a test box with a solar energy supply in order to classify wild animals based on AI methods using a combination (cross-modal learning) of radar, IR camera and IR array. The collected data will also be used to develop artificial intelligence-based data processing in order to recognize and classify wild animals such as hunting game using movement patterns. In addition to basic signal processing and the development of self-learning algorithms, the radar data must be reduced to significant features in particular. The aim is to achieve an overall object recognition accuracy of at least 95%. The data processing algorithms based on the test data obtained are linked to the prototype to be developed and optimized. A test is carried out in relevant environments such as wildlife bridges and game reserves.

FindMine Metal (FMM), FindMine Radar 2 (FMR2) and FindMine fk

Project manager: Prof. Dr.-Ing. Thomas Walter
Project duration: 01.12.2019 - 30.09.2024
Funded by: Stiftung
Project description:
FindMine Metal (FMM) is the second phase of FindMine, a joint project to detect landmines using drones. In the second phase, THU is developing metal detectors and associated signal processing and applying this system to a UAV being developed by ETH Zurich. The project is funded by the Urs Endress Foundation. Partners are the University of Ulm, ETH Zurich and the Foundation.
Based on the work from FindMine Metal, the metal detector for detecting landmines is to be further developed. In particular, the attachment to a UAV is to be optimized. Furthermore, a classification capability for different types of mines and debris is to be implemented.
FindMine Radar II (FMR2) is concerned with the UAV-supported detection of landmines. In FMR2, a ground-guided radar platform is being developed with which so-called AB scans are carried out (downlooking radar). The influences of vegetation and soil properties are being investigated and adapted signal processing is being developed. Furthermore, sensor data fusion with the metal detector is to be carried out in order to map and classify inhomogeneities in the ground.
In FindMine fk, a fast and efficient SAR algorithm for radar-based mine detection with UAV platforms is to be developed and implemented. The so-called Stolt fk migration allows real-time processing and is therefore significantly faster than back projection.

RADar makes SAFE: Radar makes safe distance tangible for bicycles

Project leader: Prof. Dr.-Ing. Thomas Walter, Prof. Michael Schlick
Project duration: 01.02.2022 - 31.12.2022
Funding provider: Federal Government - BMBF
Program name: University competition for the Science Year 2022 - In demand! Science in Dialogue (WiD)
Project description:
The aim of the project is to make cycling safer as a sustainable form of mobility in Ulm and to take interested citizens on a journey through the research cycle. A radar sensor is used, which citizens can attach to their bicycles. This measures both the distance and the speed of vehicles passing (often too closely). In this way, dangerous spots for cyclists in Ulm can be identified and documented with data.

Innovative millimeter-wave sensor technology for industrial applications (MikroSens)

Project leader: Prof. Dr.-Thomas Walter, Prof. Dr. rer. nat. Hubert Mantz
Project duration: 01.01.2019 - 31.12.2020
Funding provider: State - MWK / EFRE
Program name: Centers for Applied Research at Universities (ZAFH)
Project description:
ZAFH MikroSens is a joint project led by Ulm University of Applied Sciences with the partners Heilbronn University of Applied Sciences, Pforzheim University of Applied Sciences and Ulm University, which has been funded by both the State of Baden-Württemberg (MWK) and the EU (EFRE) since the end of 2015.
The network has set itself the goal of transferring radar sensor technology from the automotive sector to industrial applications. For example, sensors have been and are being developed at THU that can safeguard the environment of robots in the field of automation technology or classify the surface quality of roads. With the entry into terahertz sensor technology (frequency range greater than 100 GHz), a significant miniaturization of the sensors and thus also of the systems was achieved.
In addition to automation technology, other fields of application such as agriculture, environmental sensors, media sensors and medical technology were also addressed in MikroSens. An image film was created for the project, which reflects the diversity of applications for radar sensors (YouTube link). One focus of the joint project is the cooperation with medium-sized companies. In the context of MikroSens, collaborations have been initiated with more than 20 companies from Baden-Württemberg.
In addition to terahertz sensor technology, other innovative technologies are being used in the MikroSens joint project. For example, an FPGA-based backend was developed at Pforzheim University, which enables fast range Doppler maps and real-time processing using high-resolution chirp sequence modulation. At Heilbronn University of Applied Sciences, radar signal processing with neural networks (CNN) was implemented to classify the boiling process in order to prevent milk from boiling over. The University of Ulm has developed dielectric waveguides that not only conduct electromagnetic waves, but can also radiate (and receive) them in a directed manner through vaporized antenna elements. With this concept, robot arms can be wrapped around to quickly detect a potentially dangerous approach by humans.
A special highlight was a THU exhibit at the BUGA 2019 in Heilbronn. Radar-based gesture recognition was used to conduct the Württemberg Chamber Orchestra Heilbronn (including the European anthem from Beethoven's 9th Symphony). The video created for this purpose and the music played at the same tempo as the visitors conducted. However, the orchestra paid attention to quality: anyone who conducted irregularly was first given a polite warning. In the event of repetition, the orchestra stood up and left the stage (with appropriate remarks and gestures). The exhibit was tested by more than 10,000 visitors at the BUGA.

Safe and sustainable mobility for urban and rural regions through intelligent road infrastructure (SALUS)

Project manager: Prof. Dr.-Ing. Thomas Walter, Prof. Dr. rer. nat. Hubert Mantz
Project duration: 01.07.2018 - 30.07.2020
Funded by: State - MWK / EFRE
Program name: Promotion of technology transfer between universities of applied sciences (HAW) and small and medium-sized enterprises (SMEs) - HAW-KMU TT program
Project description:
Cyclists and deer have something in common: they are frequently involved in traffic accidents. On average, nine cyclists have an accident on German roads every hour. A deer accident occurs every two minutes. A research team at Ulm University of Applied Sciences led by Prof. Hubert Mantz is trying to prevent such accidents.
The aim of the project "Safe and sustainable mobility for rural and urban regions through intelligent road infrastructure" (SALUS) is to develop an intelligent road infrastructure. It is designed to detect wild animals and cyclists using radar sensors and then issue a warning to other road users. A key challenge of the research project is to develop a suitable radar sensor system using innovative and highly integrated millimeter wave technology that enables the clear classification of cyclists and wild animals. New machine learning methods will be used to evaluate the sensor data. THU is conducting research on this project together with Heilbronn University, business partners and the Baden-Württemberg Forest Research Institute.

Conformal and multistatic MIMO radar configurations for all-round visibility for automated driving (KoRRunD)

Project manager: Prof. Dr.-Ing. Thomas Walter
Project duration: 01.01.2017 - 30.04.2020
Funded by: Federal Government - BMBF
Program name: Elevate
Project description:
KoRRunD is a sub-project of the BMBF Elevate cluster. The aim is to further develop radar-based driver assistance systems for automated driving. In particular, cooperative radar sensors are being developed and validated to enable all-round visibility around the vehicle. The task of Ulm University of Applied Sciences is the development and validation of a target simulator for complex targets, with which the classification capability of radar sensors can be evaluated.

Driver assistance systems for bicycles in passenger and freight transportation (InBiDAS)

Project manager: Prof. Dr. Hubert Mantz
Project duration: 01.07.2017 - 30.06.2019
Funded by: Federal Government - BMBF
Program name: Funding of an ideas competition for the establishment and expansion of innovative R&D networks with partners in countries bordering the Danube
Project description:
The project aims to develop, implement and test various driver assistance systems for bicycles (for passenger and freight transport). For example, the use of radar sensors to detect other road users is conceivable; differentiating between trucks, cars and other bicycles, for example, is crucial here. The radar sensors can also detect road conditions and thus distinguish dry from wet or slippery surfaces. In an ageing society (which will increasingly be riding electric bikes in the future), this can lead to a significant increase in safety in combination with other sensors. Furthermore, innovative drive and energy concepts for cargo bikes are to be designed and implemented in a study in cooperation with universities and companies along the Danube. In accordance with the know-how, experience and synergies of the network partners designated to date, the scientific research and application projects will be assigned to the areas of need of intelligent mobility and sustainable management and energy of the German government's high-tech strategy. The aim of the first phase is to network German scientists and companies with partners from Serbia and Hungary. In the second phase, a joint research proposal will be developed. With InBiDAS, a network of universities and SMEs from countries along the Danube is to be established for the joint development of an EU application or similar measures.