The KARLI COnsortium
The partners of the consortium at a glance
The development of AI know-how is significantly accelerated by the ambitious project goals and the cooperation in the consortium.
The KARLI project involves various teams from the central innovation organization HEAT (Holistic Engineering and Technologies) with specialist expertise in the areas of data-driven algorithm development, machine learning, driver monitoring and human-machine interaction. The project is being carried out at the Babenhausen and Frankfurt am Main sites.
Continental is also contributing near-series sensor technology and a demonstration vehicle to the project. Continental can draw on findings and experience from past and current funded projects such as ADAS&ME, -@City and Ko-HAF.
The team brings extensive prior experience in the areas of motion sickness and possible countermeasures, as well as extensive expertise in the design, implementation, and statistical analysis of human factors studies.
In addition, Ford is contributing its own test vehicle, which already has extensive measurement technology and will be expanded to meet KARLI-specific technical requirements in its own workshops as part of the project. Among other things, Ford is contributing relevant experience from the publicly funded Interactive, InMotion, Alfasy, L3-Pilot and Hi-Drive projects.
The departments involved in KARLI are Development Innovations User Interface/User Experience, Innovation Management, Product and Test Setup and Innovation Center. Among other things, the departments contribute extensive know-how along the entire impact chain of the implementation of AI-based systems.
Audi is contributing a vehicle prototype with measurement technology to the KARLI project. AUDI AG was and is also involved in national and European research projects in the field of (highly) automated driving in close cooperation with universities and research institutions. Examples of projects include INEMAS, Ko-HAF, UR:BAN, @City and PEGASUS.
The KARLI project is supported by various Centers of Excellence (CoE) at INVENSITY. The Invensity Technology Hub, CoE Data Systems and CoE Safety Management contribute their technical expertise regarding AI, data-driven development, data management and safety management.
INVENSITY will provide the cloud-based data infrastructure for the project and draws on years of experience in both safety and systems engineering to guide the integration of sensors and connectivity to the data infrastructure in KARLI.
In the KARLI project, the research department of semvox will apply competencies in the area of dialog management and affective interaction with the use of rule- and statistics-based methods, on the basis of its own geni:OS interaction platform, complemented with speech technologies (ASR,TTS).
In several research (e.g. AutoAkzept, F-RELACS) and customer projects, TWT is developing empathic assistants for vehicles that use machine learning to detect negative user states (uncertainty, frustration) in context and reduce them through adaptive speech- and display-based MMIs.
Their expertise on AI and user-adaptive technology makes TWT an ideal link between research and industry in KARLI.
With a size of 30 employees, studiokurbos brings its expertise in the field of user-centered design process, evaluation of users, method development in the field of “digital design & development process” and effect and perception of design to the project. Further competences in the field of VR and VR interactions can be a great added value for all partners in the KARLI project.
In particular, the conception and implementation of the HMI graphics of the current Porsche generation of Macan, 911, Panamera and Cayenne should be mentioned here.
The Human Factors and Vehicle Interaction group has been working on user-centered innovations in vehicles for 30 years as contract research for OEMs and Tier1-2 suppliers as well as in numerous consortia projects (e.g. CollaborativeTEAM, SAFESPOT, UR:BAN, InCarIn, ADAS&ME, REL:AI, VORREITER, etc.).
Here, the focus is on driver state detection in driving simulators and real-world driving tests, as well as the implementation of human-related data collection for AI applications.
In the Fraunhofer IOSB department “Interactive Analysis and Diagnosis” (IAD), innovative interaction methods and assistance systems are developed. The IOSB development of Advanced Occupant Monitoring Systems is based on camera-based person perception for adaptive human-machine interactions. AI methods enable the evaluation of activities and intentions for intelligent and proactive vehicle interiors and for the interaction of vehicles with pedestrians.
In KARLI, IOSB is developing image-based AI algorithms for level-compliant behavior detection and motion sickness determination. Recognized activities and intentions are optimized for interaction with vehicle occupants, among other things.
In this project, Allround Team GmbH is primarily contributing its expertise in the field of user studies by being primarily responsible for the surveys in the Motion Sickness application.
In addition, the company is also supporting the implementation of the HMI in the test vehicle. Findings from EU-funded research projects ALFASY, interactIVe and Response3 are being utilized.
The expertise of the institute of Market and Advertising Research at the Hochschule der Medien (HdM) lies in the development and application of methods of user-centered offer design for the optimization of user experience and user behavior (e.g. regarding traffic safety).
The focus of the KARLI project is on the user-centered development and evaluation of AI-based innovations to promote level-appropriate behavior. Topic-relevant preliminary work can be found in the BMWi-funded projects TANGO and RUMBA. On the other hand, the HdM is developing and testing empirically based methods for technology assessment of ethical and social implications within the ELSI process (ELSI = Ethical, Legal and Social Implications). In collaboration with the law firm Branmatt, legal consequences of AI-based automation with regard to data protection and product liability are also being investigated.
The expertise of the Institute of Ergonomics and Technology Management (IAT) at the University of Stuttgart covers a broad spectrum of physiological sensors and data. New user-adaptive interaction concepts are designed at the institute in iterative user-oriented development steps and evaluated with potential users.
The institute contributes to KARLI above all this methodological expertise and the psychological and linguistic know-how for AI-based voice dialogs, which are indispensable for the developments of Audi and Ford.