Browse some of the best publications and dissertations by Bosch researchers

GMM Award 2021 (VDE/VDI Society for Microelectronics, Microsystems and Precision Engineering)

Paper presents a prototype of a miniaturised and integrated magnetometer prototype based on colour centres in diamond.

Best Paper Award at FSE Conference 2021

Arguably one of the main applications of the LowMC family ciphers is in the post-quantum signature scheme PICNIC. Although LowMC family ciphers have been studied from a cryptanalytic point of view before, none of these studies were directly concerned with the actual use case of this cipher in PICNIC signature scheme. Due to the design paradigm of PICNIC, an adversary trying to perform a forgery attack on the signature scheme instantiated with LowMC would have access to only a single given plaintext/ciphertext pair, i.e. an adversary would only be able to perform attacks with data complexity 1 in a known-plaintext attack scenario. This restriction makes it impossible to employ classical cryptanalysis methodologies such as differential and linear cryptanalysis. In this paper we introduce two key-recovery attacks, both in known-plaintext model and of data complexity 1 for two variants of LowMC, both instances of the LowMC cryptanalysis challenge.

Best Paper Award at EWSN Conference 2021

Recent advances in computing have allowed for the possibility to collect large amounts of data on personal activitiesand private living spaces. To address the privacy concernsof users in this environment, we propose a novel frameworkcalled PR-GAN that offers privacy-preserving mechanismusing generative adversarial networks. Given a target application, PR-GAN automatically modifies the data to hide sensitive attributes – which may be hidden and can be inferredby machine learning algorithms – while preserving the datautility in the target application. Unlike prior works, the public’s possible knowledge of the correlation between the targetapplication and sensitive attributes is built into our modeling. We formulate our problem as an optimization problem,show that an optimal solution exists and use generative adversarial networks (GAN) to create perturbations. We fur-ther show that our method provides privacy guarantees under the Pufferfish framework, an elegant generalization ofthe differential privacy that allows for the modeling of priorknowledge on data and correlations. Through experiments,we show that our method outperforms conventional methodsin effectively hiding the sensitive attributes while guaranteeing high performance in the target application, for bothproperty inferenceandtrainingpurposes. Finally, we demonstrate through further experiments that once our model learnsa privacy-preserving task, such as hiding subjects’ identity,on a group of individuals, it can perform the same task on aseparate group with minimal performance drops.

Best of Innovation Award, CES 2020

The Bosch Virtual Visor was honored as a Best of Innovation in the CES 2020 Innovation Awards. Virtual Visor links an LCD panel with a driver or occupant-monitoring camera to track the sun’s casted shadow on the driver’s face. The system uses artificial intelligence to locate the driver within the image from the driver-facing camera. It also utilizes AI to determine the landmarks on the face — including where the eyes, nose and mouth are located — so that it can identify shadows on the face. The algorithm analyzes the driver’s view, darkening only the section of the display through which light hits the driver’s eyes. The rest of the display remains transparent, no longer obscuring a large section of the driver’s field of vision.

Best of Innovation Award, CES 2020

Bosch’s new 3D display uses passive 3D technology to create a real-feeling three-dimensional effect of images and warning signals. This allows information to be captured more quickly than on conventional screens. The driver is less distracted as a result. The system for spatial representation does not require any additional technologies such as eye tracking or 3D glasses.

IEEE VAST Best Paper Award 2020

Traffic light detection is crucial for environment perception and decision-making in autonomous driving. State-of-the-art detectors are built upon deep Convolutional Neural Networks (CNN) and have exhibited promising performance. However, one looming concern with CNN based detectors is how to thoroughly evaluate the performance of accuracy and robustness before they can be deployed to autonomous vehicles. In this work, we propose a visual analytics system, VATLD, equipped with a disentangled representation learning and semantic adversarial learning, to assess, understand, and improve the accuracy and robustness of traffic light detectors in autonomous driving applications. The disentangled representation learning extracts data semantics to augment human cognition with human-friendly visual summarization, and the semantic adversarial learning efficiently exposes the interpretable robustness risks and enables minimal human interaction for actionable insights. We also demonstrate the effectiveness of various performance improvement strategies derived from actionable insights with our visual analytics system, VATLD, and illustrate the practical implications for safety-critical applications in autonomous driving.

Cooperation award science-economy of Ulm University 2020

As part of the transfer project “Do it yourself, but not alone: Companion technology for do-it-yourself assistance”, a prototype Companion system for do-it-yourself assistance is to be developed and practically tested in close cooperation between the Ulm University and the Corporate Sector Research and Advance Engineering Department of Robert Bosch GmbH. Based on the research results of the subprojects A1, A2 and B1 of the collaborative research center Transregio 62, a digital assistant is to be developed for the first time that will provide users with individual and competent support in operating power tools and carrying out DIY projects. Among other things, the assistant is able to adapt its detailed step-by-step instructions individually to the situation as well as to the abilities of the respective user. For the implementation of the respective assistance task, the system can draw on a comprehensive knowledge base in the form of formal specifications. In addition, the assistant is also intended to allow users to specify their owns DIY projects and assist them with these. These authorized specifications can then be made available on a community platform for other users.

Werner von Siemens Ring award for young scientists 2019

The award of Andreas Dörr as a young scientist of the Werner-von-Siemens-Ring Foundation is in recognition of his outstanding work for data-efficient reinforcement machine learning.

One of the great promises of machine learning (ML) and artificial intelligence (AI) is to give machines the ability to autonomously learn new behavior in complex and unknown environments. Results of this research direction, in particular of the discipline of reinforcement learning (RL), are however still predominantly limited to very specialized, simulatable and well-understood problem areas. The work of Andreas Dorr, on the other hand, addresses problems such as: data efficiency, robustness and learning in the presence of erroneous and incomplete perception of the environment. These problems typically arise as soon as machine systems are asked to operate in our unstructured, everyday environments. Core topics of his work are the learning of probabilistic, predictive models, which allow the machine to derive optimal decisions for future actions from past experiences.

Picture: Stefanie Winkler (München)

Stiftung Werner-von-Siemens-Ring

Werner von Siemens Ring award for young scientists 2019

The award of Gerhard Kurz as a young scientist of the Werner-von Siemens-Ring Foundation is in recognition of his outstanding contributions to simultaneous localization and mapping for mobile robots.

Modern mobile robots have to find their way in unknown environments and interact safely with their environment. For this purpose, it is of great importance that they create a map of the surroundings and locate themselves on the basis of this map. This is called Simultaneous Localization and Mapping ( SLAM). In this process, a robot detects its environment and its own movement within this environment with the help of various sensors. Information from all sensors is combined to provide the most accurate and reliable results possible. SLAM has applications in many fields, from vacuum cleaning robots to logistics robots, from autonomous vehicles to flying robots. There are many challenges to overcome, such as how to accurately represent the uncertainties that arise or how to adapt the map to a changing environment.

Picture: Stefanie Winkler (München)

Stiftung Werner-von-Siemens-Ring

Best Paper Award, IEEE International Conference on Mechatronics 2019

Future automotive technologies become more and more autonomous and connected. This trend requires a rethinking of validation processes due to the amount of test kilometers needed. To be able to test automated and connected functions in many different traffic scenarios, virtual and mixed real-virtual prototypes will be used. Moreover, due to the complexity of such systems, cross-company cooperation is increasing and demands for common prototypes. Spatially distributed prototypes simplify and enhance cross-company collaboration due to faster provisioning of models and better IP protection. However, the setup of such prototypes is very time consuming due to the high integration effort. Here it is shown that the integration effort of spatially distributed prototypes can be massively reduced by using the Distributed Co-Simulation Protocol (DCP). A demonstrator consisting of a small scale test bed located in Graz and a co-simulation containing Bosch driving functions located in Renningen is presented. The demonstrated integration workflow as well as an analysis of the communication challenges of the coupling can be transferred to any other coupling of this kind.

Werner von Siemens Ring award for young scientists in engineering and natural sciences 2015

The award of Tobias Diekhans as a young scientist of the Werner-von-Siemens-Ring Foundation is in recognition of his groundbreaking scientific achievements in inductive energy transmission for electric vehicles.

A wireless charging system offers the potential to make the “refuelling” of future electric vehicles as convenient as we are used to from the combustion engine today. The core of the energy transmission is formed by two inductively coupled coils: one is integrated into the parking space and the second is mounted under the vehicle. Further power electronics components first convert the low-frequency mains voltage into a high-frequency alternating voltage and then back into a direct voltage in the vehicle. The power electronics topology with regulation on both sides represents an extremely promising approach to meeting the high requirements in terms of installation space, efficiency and costs.

Werner von Siemens Ring award for young scientists in engineering and natural sciences 2013

The award of Andreas Burghardt as a young scientist of the Werner-von-Siemens-Ring Foundation is in recognition of his scientific achievements in the field of powder injection molding, which have contributed significantly to the improvement of the microscopic understanding of the manageability of particularly nanoscale materials in the manufacturing process.

Powder-technological shaping processes can be used to manufacture complex components from metal or ceramics. These processes require dispersions with a solids content of up to 70% by volume, which can only be achieved by overcoming the attractive Van der Waals forces with the aid of dispersants. When nanopowders are used, the attraction is so strong that conventional dispersants are no longer effective. A highly viscous thermoplastic as a medium effectively prevents the agglomeration of the nanoparticles, but the polymer molecules adsorb completely on the nanoparticles, so that the dispersion loses its plasticity. Using a new approach, the adhesion energy of the polymer molecules can be reduced to such an extent that even dispersions with nanopowders remain processable.

Thomas Gessner Award 2019

At Smart Systems Integration Conference and Exhibition SSI 22019 for the first time the "Thomas Gessner Award" has been granted. The winner is Dr. Christian Huber, Robert Bosch GmbH Renningen. He obtained the prize for his PhD "Micromechanical Tunable Fabry-Pérot Interferometers with Membrane Bragg Mirrors Based on Silicon/Silicon Carbonitride". Especially, the new MEMS FPI design with two released membrane mirrors and the demonstration of the feasibility of his concept need to be mentioned.

Techtextil Innovation Award 2019 Category 'New Technology'

Robert Bosch GmbH and H. Stoll AG were presented with the Techtextil Innovation Award for a knitted sensor glove. The seamless, 3D flat-knitted glove is made of sensor yarn and offers the wearer sensory and control functions on all fingers, e.g., for operating interfaces in man-machine interaction, for movement control in relation to augmented and virtual-reality applications or rehabilitation purposes in a medical setting. The glove provides the same level of wearing comfort as a conventional glove.

Best Paper Award, IEEE Conference on Visual Analytics Science and Technology (VAST) 2018

Consider a multi-dimensional spatio-temporal (ST) dataset where each entry is a numerical measure defined by the corresponding temporal, spatial and other domain-specific dimensions. A typical approach to explore such data utilizes interactive visualizations with multiple coordinated views. Each view displays the aggregated measures along one or two dimensions. By brushing on the views, analysts can obtain detailed information. However, this approach often cannot provide sufficient guidance for analysts to identify patterns hidden within subsets of data. Without a priori hypotheses, analysts need to manually select and iterate through different slices to search for patterns, which can be a tedious and lengthy process. In this work, we model multidimensional ST data as tensors and propose a novel piecewise rank-one tensor decomposition algorithm which supports automatically slicing the data into homogeneous partitions and extracting the latent patterns in each partition for comparison and visual summarization. The algorithm optimizes a quantitative measure about how faithfully the extracted patterns visually represent the original data. Based on the algorithm we further propose a visual analytics framework that supports a top-down, progressive partitioning workflow for level-of-detail multidimensional ST data exploration. We demonstrate the general applicability and effectiveness of our technique on three datasets from different application domains: regional sales trend analysis, customer traffic analysis in department stores, and taxi trip analysis with origin-destination (OD) data. We further interview domain experts to verify the usability of the prototype.

Blue Shield Technology Innovation Award by China International Association for Promotion of Science and Technology 2020

The Secure Product Fingerprint is a patented solution developed by the Bosch research team in China. Its primary function is to ensure product authenticity. Based on computer vision and AI technologies, it takes a “fingerprint” of a product and stores it in the cloud. This unique “fingerprint” is either an assigned QR code enhanced with micro-features (“Artificial Fingerprint”) or a snapshot of one particular area of a product’s surface (“Natural Fingerprint”). Using a smartphone, customers can then scan the QR code label or the designated area of the product itself and verify the product’s authenticity by comparing it to the sample stored in the cloud. This can be done via a dedicated app, WeChat, or a standard QR code scanner.

Best Paper Award 2020

The use of neural networks in perception pipelines of autonomous systems such as autonomous driving is indispensable due to their outstanding performance. But, at the same time their complexity poses a challenge with respect to safety. An important question in this regard is how to substantiate test sufficiency for such a function. One approach from software testing literature is that of coverage metrics. Similar notions of coverage, called neuron coverage, have been proposed for deep neural networks and try to assess to what extent test input activates neurons in a network. Still, the correspondence between high neuron coverage and safety-related network qualities remains elusive. Potentially, a high coverage could imply sufficiency of test data. In this paper, we argue that the coverage metrics as discussed in the current literature do not satisfy these high expectations and present a line of experiments from the field of computer vision to prove this claim.

Eduard-Rhein Technology Award 2019

Andrea Urban and Franz Lärmer received the technology award of the EDUARD-RHEIN-STIFTUNG for their invention of the deep reactive-ion etching – also known as Bosch process. It is a key process for the production of semiconductor sensors (i.e. MEMS sensors). This prize puts them into the row of an exclusive community of very prestigious winners, i. e. the inventors of the World Wide Web or the MP3-audio-coding.

Werner von Siemens Ring award for young scientists in engineering and natural sciences 2017

The award of Julia Vinogradska as a young scientist of the Werner-von-Siemens-Ring Foundation is in recognition of her outstanding achievements for the further development of theoretical foundations for the design of reliable systems with artificial intelligence.

A number of breakthroughs have been made in artificial intelligence in recent years. This technical progress makes it possible to address control engineering issues with new approaches. From system identification to controller synthesis to stability verification, artificial intelligence can help solve problems as efficiently and cost-effectively as possible while minimizing manual effort.

European Inventor Award Category 'Industry' 2007

When German inventors Dr. Franz Lärmer and Andrea Urban created the Bosch Process by adapting micro-process techniques, they opened up a new world of affordable car safety devices.

Lärmer and Urban developed a method for manufacturing high-precision silicon sensors using plasma technology while working in sensor and process development at Robert Bosch Corporate Research. The purpose of their invention was to etch deep microstructures with vertical sidewalls into silicon wafers at high speed and with great accuracy.

The invention concerns a method for the anisotropic etching of features defined by an etching mask (preferably recesses with precisely defined sides in silicon) produced using a plasma etching technique.

The etching process has been around for some 30 years in the world of MEMS. A common problem with etching deep into silicon was that the process also etched sideways, eating into the wall of the structure being created and resulting in distorted walls.

The inventors' method overcame the problem by using a fluorocarbon-based plasma to deposit an etch-resistant layer before taking a subsequent etch step. The technique, nowadays simply referred to as the Bosch Process, allows deep etching in a well-defined manner.

The technology revolutionised MEMS at the time and is currently used worldwide to manufacture silicon MEMS.

German Future Prize 2013

Light is a versatile tool. In manufacturing applications, laser beams can replace conventional mechanical machine tools for boring, cutting or even welding. They make the machining of materials fast, cost efficient and, above all, precise. But how can the highest precision with maximum productivity be achieved with ultrashort laser pulses?

Dr. Jens König, Dr. Dirk Sutter and Prof. Dr. Stefan Nolte have mastered this challenge: they developed ultrashort laser pulses for industrial applications. These lasers emit light in the form of high-energy laser pulses no more than a billionth of a second in duration. They have thus paved the way for the very precise manufacture of minute structures – and for the industrial production of innovative products. Jens König is group leader for laser material processing and joining technology in the Research and Preliminary Development division of Robert Bosch GmbH in Schwieberdingen. Dirk Sutter heads Research and Development of ultrashort laser pulses at TRUMPF Laser GmbH + Co. KG in Schramberg. Stefan Nolte is a professor at the Institute of Applied Physics at the Friedrich Schiller University of Jena as well as at the Fraunhofer Institute of Applied Optics and Precision Mechanics in Jena.