FlexCAR

1. the vision

The vision of the FlexCAR collaborative project is to apply this development to future mobility concepts. Suppliers, as well as new market participants such as startups and developer communities, are invited to provide services and components throughout the entire vehicle lifecycle—either to platform operators (OEMs) or directly to end customers. To this end, hardware and software interfaces must be defined. They must be as easy to use as possible (“plug & play”) while still meeting the highest cybersecurity standards.

2. the objective

The aim of the FlexCAR project is to develop a cyber-physical vehicle platform with open hardware and software interfaces. The implementation of self-reflecting, self-describing and service-oriented components on this platform makes it possible to design a vehicle that can be updated and upgraded over its entire life cycle in both the software and hardware areas, including complete, automatic reconfigurability and safeguarding of the overall system. In order to demonstrate the feasibility of the concept and to show how the automobile of tomorrow will be created and designed, the FlexCAR joint project is examining the open interface issues over the course of the project using typical vehicle scopes. These vehicle scopes include the development of individual, future-oriented components in an exemplary, scaled-down innovation network and their use in a research and experimentation platform.

The FlexCAR project focuses on four thematic areas:

Subject area 1 - The "Rolling Chassis" is developing an electrically powered research platform controlled via 5G with accessible and open software and hardware. The platform serves as a research and experimentation platform for the development of individual technologies in this project, independent of competition and model series. This platform is supplemented by a base plate, which creates an interface to the interior and is available to TB 2 for various use cases and interior studies.

The focus of Topic Area 2—“Interior and Cyber-Physical Access”— builds on the topics covered in Topic Area 1. Here, interior-related use cases are examined, with a focus on the areas of “Working during autonomous driving” and “Relaxing during autonomous driving.” For example, if a customer books a ride via an app, this request is sent and processed via 5G. A vehicle is assigned according to the customer’s requirements, and the customer is provided with details regarding the location and time. The vehicle arrives at the agreed-upon location at the agreed-upon time and looks out for the customer. Contact with the vehicle can be established via outward-facing cameras. Access is granted only to authorized persons, and the ride can be billed accordingly. The future interaction between passengers and autonomous vehicles is therefore also a key focus during the ride.

In topic area 3 - "Flexible product and production concepts", the FlexCAR production and assembly concept was developed in close cooperation with the Fluide Production(FluPro) project, among others. Drive and energy modules are the focus of production engineering considerations here; for example, the issues of suitable sheet metal or structural profiling within the chassis. With the help of standardized interfaces, the rolling chassis is part of the cyber-physical production system and can be used as a further use case during production, for example as a driverless transport system, thus replacing current transport and conveyor technology in the factory.

Subject area 4 - "Open development process for modular vehicles" deals with the implementation of one of the key criteria of FlexCAR: the open development process. The aim is to develop and implement open interfaces for hardware and software, allowing new business models to be established. FlexCAR is therefore much more than just a new vehicle concept. It offers a research platform for the central development and production of modular vehicle systems.

4. the results

4.1 Drive and alternating storage systems

Motivation
As part of the FlexCAR research project, a mobile, updateable and upgradeable platform was built as a rolling chassis, which essentially consists of three individual modules. These three individual modules, which are divided into two identical drive modules (DRIVE modules) at the front and rear and a storage module (energy module) in the middle of the chassis, form the basis for the rolling chassis. Pre-equipped with drive components and E/E scopes in the DRIVE modules, these can then be joined together with the Energy module. Together with the integrated energy storage unit, the energy module forms the centerpiece of the chassis. Different energy storage systems (battery or fuel cell) and associated components can be integrated into it. By largely standardizing the interfaces and contact points, the DRIVE modules and Energy modules can be combined to create a drive-flexible platform.

Status
The different energy storage systems are integrated using customized support structures. The interfaces between the individual modules for electrical and electronic components are largely standardized via connector systems and connections. Findings from the " Wire HarnessInnovation Initiative", an open cooperation network of actively involved industry representatives working on various pre-competitive topics as part of the ARENA2036 research campus, were incorporated into this process.

4.2 Floor module structure, interior set-up and use cases

Motivation
Another research focus of the FlexCAR project was the design, demonstration and evaluation of new interior concepts for the autonomous mobility of the future. Use case work plays a particularly important role in making the best possible use of driving time in automated or autonomous vehicles.

Stand
The hardware interface between the rolling chassis and the interior theme area is the interior module. This has a self-supporting frame construction into which seat rails with additional side rails have been integrated. This rail system can be used to create various seat constellations and positions, as well as to attach different interior components and position them independently of the seats. The "mobile work 2036" user study was designed to investigate different technologies for the consumption and processing of digital content. To this end, the interior module was equipped with two seats and two interactive digital work surfaces. The first interactive surface consists of two HPL panels with a touch-sensitive film laminated between them. With the help of three projectors, which are realistically suitable for installation in the car due to their design, an interactive interior surface was created.

4.3 Exterior and interior concept developments

Motivation
The shift in vehicle usage models toward autonomous driving and shared mobility is directly linked to changing mobility needs and usage patterns, and thus places greater demands on the flexibility and adaptability of tomorrow’s vehicles. The FlexCAR research project focused on creating a unified research platform that can help reduce both development time and costs. Modern, cyber-physical approaches make it possible to incorporate production aspects and product themes into customer-centric exterior and interior designs that reflect various use cases.

Stand
The focus was on two different main use cases that demonstrate the flexibility of the rolling chassis. The Peoplemover Use Case focuses on the transportation of 4 passengers using the shuttle as a combination of "Working during driving autonomous" and "Relaxing during driving autonomous". The Luxury Car Use Concept serves to illustrate and investigate future autonomous high-class mobility, where the focus is on "Relaxing while driving autonomously". These use cases require different exterior and interior designs (see Fig. 5 and Fig. 6), which we created and examined in virtual reality. Specific features were built in hardware, but the majority were represented digitally as 3D models and examined cyberphysically. Specific use cases of this research were the testing of different boarding concepts, as well as communication and interaction in the interior between the passenger and the autonomous shuttle.

4.4 Cyberphysical visualization technology

Motivation
As part of FlexCAR, the Hochschule der Medien focused on making autonomous vehicles and the vehicle interior of the future tangible. The project looked at driverless autonomous driving at SAE level 5 with a focus on first-time users and early adopters. The focus here was on creating transparency between the passenger and the autonomous vehicle. By creating AI transparency and breaking down the actions of the autonomous vehicle using innovative human-machine interfaces, the aim was to create a positive user experience for end users during autonomous driving.

Status
The implementation of tracking real vehicle elements and their synchronization with a virtual 3D environment is well advanced. A virtual autonomous drive through a simulated city with various parameters (e.g. weather, time of day, traffic volume, etc.) has been implemented. The next steps included finalizing the interior and exterior of the autonomous shuttle as well as test drives to iteratively adapt the prototype. Furthermore, performance optimizations and the expansion of the current status to include elements from project partners (e.g. luggage systems, mobile working or boarding concepts) were carried out.

4.5 New profile structures for the FlexCAR chassis

Motivation
The rolling chassis developed in the FlexCAR project is a mobile, upgradeable and modular vehicle platform that can be powered purely by electricity or a fuel cell. The current FlexCAR prototype follows a highly integrative conceptual approach, which was optimized in a second step through flexible, demand-oriented production concepts. The focus here was on the further development of the energy storage module (energy module), which formed the framework structure for the high-voltage storage system and for the fuel cell with hydrogen tanks. New requirements resulting from the structural change in the automotive industry towards sustainable, electric mobility and from shorter development and product life cycles were taken into account. The profile structures developed here should be as weight-optimized as possible, easily recyclable, designed to meet requirements and easily adaptable.

As of
The simulation-based optimization of the energy module made of aluminum extruded profiles with regard to a modular design while simultaneously meeting crash requirements relevant to type approval—in particular the side pole test—showed that incorporating local reinforcement structures can reduce the weight of the sill while improving crash performance. The intelligent combination of different aluminum alloys not only ensures that the components possess the required properties—balancing ductility and extrudability—but also guarantees that the sill profile maintains system integrity even under severe deformation. The two insert profiles in the side member, which were integrated into the side member in a modular fashion depending on the configuration, consist of a high-performance HSA6® aluminum alloy with excellent recyclability, thereby making a significant contribution to weight reduction and sustainability.

To meet crash requirements even in an all-sheet-metal design, the crash-relevant profiles—composed of several bent sheets—were joined using laser beam welding to form a multi-chamber profile. By utilizing TRUMPF’s innovative multifocus technology, the formation of pores in the weld seam was largely prevented, enabling gas-tight weld seams. In some cases, the sheet metal panels were positioned using snap-fit connections. This enables flexible joining with minimal fixturing. The elimination of the flange also reduces weight.

In conventional manufacturing, the corner joint is typically produced as a separate component—such as a cast part—and usually requires extensive rework due to fit issues. In this project, the corner joint was created through intelligent nesting of the sheet metal profiles using an innovative bonding technique. To achieve this, an LMD weld—whose height corresponds exactly to the gap dimension between the profiles—was applied to one of the profiles to be bonded, and the profiles were then assembled. This approach not only compensated for tolerances in the gap dimension but also ensured a consistent gap dimension throughout. The consistent gap dimension and precise positioning thus enable reliable adhesive injection and a secure bond between the profiles without critical heat input.

4.6 Sensor integration on the FlexCAR

Motivation
The integration of sensors in the rolling chassis of the FlexCAR project was intended to create a vehicle that can be updated and upgraded. This includes the sensor technology for autonomous driving as well as the use of sensors for structural monitoring and the integration of additional functions.

Status
The vehicle's internal wheel speed and steering angle sensors and the IMU were used in an approach for estimating driving conditions and localization. The speed, alignment and position prediction from an inertial navigation system was combined in an Unscented Kalman Filter (UKF) with the observations from a wheel speed-based odometry model and a gravity-based position calculation. This system could be supplemented by position observation from 5G positioning or infrastructure camera-based localization. Lidar and camera are used in FlexCAR for environment modelling, localizing the vehicle and detecting objects in the immediate vicinity.

4.7 Implementation of printed control element with integrated RFID component recognition

Motivation
In order to be able to manufacture electrical components such as a control element for a vehicle cost-effectively, they usually have to be produced in large quantities. At the same time, however, these components should also be able to be integrated into the corresponding vehicle interior in the most visually appealing way possible in terms of color and design, which results in a correspondingly high number of variants. With the help of additive manufacturing processes, the design, surfaces and colors of components can be freely designed during production on the same machine. It is also possible to freely position printed conductor paths in the component and integrate printed sensor elements. Wireless verifiability is desirable to ensure that the correct component is and has been installed in the respective vehicle with a high number of component variants.

As of
Using inkjet technology, an additively manufactured 2D demonstrator of a control element (dimensions 100x50x1 mm³) with pushbutton and slider functions was produced. It consists of a semi-transparent black plastic, printed circuit traces and sensor elements, and an electrical connection to a 16-pin ribbon cable. Furthermore, by selecting appropriate printing parameters and strategies, the surface was visually and haptically adjusted to such an extent that the position of the button is visible and can also be felt with a finger. RFID tags (dimensions: 57.1 × 5.95 × 1.3 mm³) were successfully integrated into inkjet-printed 3D objects and could be identified remotely using an RFID reader. Through a user interface developed as part of the project, the data on the RFID tag can be displayed and rewritten as needed.

4.8 Integrated onboard and offboard communication via 5G in the ARENA2036-Architecture of the onboard and offboard network

Motivation
Requirements for future vehicles are increasingly shifting from hardware to flexible software instances in order to be able to react to changing demands on computing resources through functional adaptations, updates or hardware failures. Flexibility in development plays just as important a role as adaptations via updates after the vehicle roll-out. The improvement of existing algorithms, the exchange with environmental sensors or the integration of external sensors beyond the system's own limits require a high degree of flexibility, which can be achieved by means of an open development platform and software-oriented architecture. Updates are already carried out over-the-air via remote maintenance without the need for workshop visits. In addition to control functions, remote access also enables the sharing of all onboard data for visualizations. At the same time, communication with external devices is the minimum requirement for Vehicle-2-everything connectivity (V2X). Integrated but flexible onboard and offboard communication via 5G in the indoor ARENA2036 environment was necessary for the FlexCAR project.

As of
The Rolling Chassis’s communication architecture encompasses both internal (onboard) and external (offboard) communication. Onboard communication refers to the exchange of data between all devices and instances that are mounted and installed on the Rolling Chassis. Information is exchanged via Ethernet. By using a service-oriented architecture (SOA), the software is no longer tied to specific hardware and can be executed flexibly within the system. The SOA is implemented in the Rolling Chassis via the Data Distribution Service (DDS) middleware, including RTI connext. The service interfaces are standardized in the .idl format and are exchanged along with additional information regarding Quality of Service (QoS). As a result, onboard communication is interoperable across software instances, programming languages, and various operating systems. This ensures adaptive compatibility with the Robot Operating System 2 (ROS2) or the AUTOSAR Adaptive Platform. For offboard communication, a 5G connection to the indoor ARENA2036 production network was established. Communication within the ARENA2036 network took place via the Message Queuing Telemetry Transport (MQTT) protocol.

Both DDS and MQTT work according to the publish-subscribe principle, but on the basis of different protocols. The protocols therefore had to be translated. To ensure full compatibility between the onboard and offboard systems, a DDS2MQTT bridge was developed by the IAT at the University of Stuttgart in collaboration with the Fraunhofer IAO. It can be used to route any messages from and to the vehicle. In addition to DDS and MQTT, the communication architecture also supports the User Datagram Protocol (UDP) for connecting embedded hardware. This made it possible, for example, to implement communication with the control unit for basic driving functions. A separate software bridge was also developed for this. The driving functions, sensors and actuators, data processing and utilization are all controlled internally by other specially developed software instances. The necessary computing power is provided by four Nvidia Jetson AGX Xavier modules. Due to the chosen architecture, the entire network communication of the rolling chassis functions as a closed unit both within the vehicle and via the indoor ARENA2036 network.

The Rolling Chassis communicates with the ARENA2036 indoor network via the ARENA2036’s 5G Standalone (5G SA) system installed by Nokia. Using two Pico Remote Radio Heads that transmit over the 5G radio interface, radio coverage of 4,000 m² can thus be achieved. An additional micro base station was installed outdoors in the ARENA2036 parking lot to enable seamless handoffs between radio cells. The core network was upgraded to a broadband 5G core network. A 10 Gbit fiber-optic line provides the necessary bandwidth. The 5G network operates with a downlink speed of over 1 Gbit/s and an uplink speed of 250 Mbit/s, with optimized round-trip latencies ranging from 8 to 15 ms.

Infrastructure-based sensor technology for position and obstacle determination plays an important role in the FlexCAR project, especially when operating indoors where no GPS signal is available. The position and location of the rolling chassis are determined outside the vehicle and transmitted to the vehicle via the 5G air interfaces.

Four IP cameras installed in the hall of the ARENA2036 were used for localization by means of a camera system. Their video data is transmitted to a central control computer via 5G. Perception algorithms are implemented there to determine the position and orientation of the rolling chassis as well as the position of any obstacles. This is done using neural networks and inverse perspective mapping (IPM). The exact position of the obstacles is determined by a "footprint" determination using AND linking of the images from the various cameras. The detected positions and obstacles are transmitted to the rolling chassis with an accuracy of approx. 5 cm using the described communication architecture.

For the localization of mobile devices, Nokia Bell Labs has also set up a prototype system in the ARENA2036 hall that goes beyond the 5G network. These are measuring receivers that receive 5G uplink signals and transmit them to a server. On the server, the position of the mobile device is estimated from the measurements using special algorithms. The 5G positioning data is returned to the vehicle in real time.

The concept of the open development platform and its implications were examined in DXC's FlexCAR project. The software-defined approach of an open development platform promotes open source and crowd collaboration with the associated inherent innovation potential. At the same time, fundamental changes are taking place in digital lifecycle management and the software supply chain. New processes such as DevOps have been integrated into the vehicle industry, along with over-the-air (OTA) updates. To demonstrate the technical feasibility, DXC designed a pilot application for OTA software updates using Excelfore's eSync technology. As an example of embedded functions, DXC selected the object recognition system used on the rolling chassis. A new version of the object recognition software was used to successfully test an improvement in ADAS functionality through the OTA software update. In collaboration with the company Asvin, DXC investigated the integrity of cyber security in the ecosystem of multi-level supplier and OEM software chains. Using an automated end-to-end process chain with regard to development, provision/release and installation and the Asvin blockchain documentation, it was possible to demonstrate how a secure, auditable software supply chain can be implemented.

5. The starting points / follow-up projects

The selected focus topics show that these new innovative approaches and methods can be used to design future forms of mobility in a targeted, fast and efficient manner with regard to automated or autonomous driving using the FlexCAR research and experimentation platform. The developed software-oriented architectures offer a flexible assignment of programs to the FlexCAR hardware and thus raise the modularity of the FlexCAR concept to a new level, which, together with broadband communication, enables complete data exchange for the use of edge or cloud computing. The open FlexCAR research platform created, with its new removable storage and production concepts, offers an ideal space for further research and development work, independent of model series and competition, with the embedding of VR/AR technology.

In Phase III, from 01/2024, selected aspects will be further researched in the follow-up project CARpulse in the Mobility 2036 funding priority on the Arena 2036 campus. These will deal with the further development of the software and hardware of this research and experimentation platform using selected demonstrator components, the visualization of selected issues using AR/MR/XR, as well as aspects of human crash modelling in autonomous driving.