We have recently completed a first round of research on In-Vehicle Environment (IVE) software stacks. This includes automotive-ready software stack which enables the execution of In-Vehicle Infotainment applications on both cluster and head unit screens. Events and data can be exchanged between screens. Multiple operation environments are supported: we have devised an AGL/Yocto + GENIVI architecture with the application environment based on HTML5 and WebGL; but also, we have an application stack entirely based on Android/Java with custom GUI development framework fully relying on OpenGL. All the aforementioned items guarantee the immersive experience for the driver.

Additionally, our IVE stacks are fully interlinked with modern Advanced Driver Assistence System (ADAS); we receive events and video feeds from ADAS systems, and we can also instruct ADASimg_20160901_112407-2 to swap to a desired algorithm or to calibrate. In the scope of our research project is also a software stack for ADAS platforms for demonstration purposes of the hardware; this stack is what we call Automotive Machine Vision (AMV) middleware, and it allows the implementation and porting of various vision-based and multi-sensor algorithms regardless of the specific SoCs used on the hardware target.


We essentially “spammed” this year’s IEEE ICCE Berlin 2016 conference with our research papers: https://edas.info/p21929


Paper1: Pilot In-Vehicle Infotainment Learning Platform based on Open Source Technologies

Aleksandar Dakic (Faculty of Technical Sciences – University of Novi Sad, Serbia); Tijana Srejic (Faculty of Technical Sciences, University of Novi Sad, Serbia); Mladen Krbanjevic (RT-RK Computer Based System, Serbia); Milan Z. Bjelica (University of Novi Sad, Serbia)
Current education institutions lack adequate tools for teaching software programming for automotive area, especially In-Vehicle Infotainment (IVI), with available tools constrained to a palette of proprietary, closed solutions. In this paper we give an overview of a pilot In-Vehicle Infotainment Learning platform (PIVILP), based on open source technologies and achievable in any educational environment with moderate cost. The developed pilot and the accompanying software libraries demonstrate the feasibility of the proposed approach during laboratory work for students at the university.


Paper 2: Automotive Cluster Graphics:Current Approaches and Possibilities

Jasmina Pesic (Faculty of Technical Sciences, Serbia); Kristina Omerovic (Faculty of Technical Sciences University of Novi Sad, Serbia); Ivana Nikolic(Faculty of Technical Sciences, University of Novi Sad, Serbia); Milan Z. Bjelica (University of Novi Sad, Serbia)
In this paper we give an overview of current approaches in the development of digital cluster units with rich graphical interface. Given that most of the presented approaches are closed source, proprietary and dedicated to a unique target environment, we discuss possibilities of utilization of open source technologies for cluster graphics development purposes. We give results of an experiment of cluster application developed using HTML5/WebGL technologies. Finally, we contrast different approaches with regard to the development time, scalability, portability, as well as the functional scope and show that utilization of HTML5 can be a viable route for future research.


Paper 3: Java-based Graphical User Interface Framework for In-Vehicle Infotainment Units with WebGL Support

Radivoje Ostojic (Faculty of Technical Sciences of Novi Sad, Serbia); Jasmina Pesic (Faculty of Technical Sciences, Serbia); Goran Stupar (University of Novi Sad & RT-RK Institute for Computer Based Systems, Serbia); Milan Z. Bjelica (University of Novi Sad, Serbia)
In-Vehicle Infotainment (IVI) applications are gaining more and more attention in the automotive industry. For application development different technologies are used, typically Linux-based, accompanied with various proprietary IDE tools. Cross-platform applications are therefore not easy to create, ramp up time to start a development project is large and require specific know-how. In this paper, we propose an application environment based on open technologies, such as HTML5 and JavaScript for IVI application development, fully utilizing WebGL for advanced graphical effects. To further facilitate the development, we created a Java-based framework which allows creating and portability of cluster and IVI applications to the proposed application environment. Evaluation shows the feasibility of our approach and adequate graphical performance.


Paper 4: Supporting Sensor Fusion in Next Generation Android In-Vehicle Infotainment Units

Kristina Omerovic (Faculty of Technical Sciences University of Novi Sad, Serbia); Joakim Janjatovic (Faculty of Technical Sciences, University of Novi Sad, Serbia); Milena Milosevic and Tomislav Maruna (RT-RK Computer Based Systems LLC, Serbia)
Each day vehicles are becoming more intelligent and users are presented with continuous enhancement of features and functions. In-Vehicle Infotainment (IVI) and Advanced Driver Assistance System (ADAS) are becoming very important for improved passenger experiences and safety within vehicles. Modern cars have multiple screens running in an environment based on multicore SoC (System on a chip), therefore it is necessary to design real time communication between ADAS and IVI. This paper proposes the communication solution between ADAS and IVI through Ethernet connection.


Paper 5: Augmented Reality Obstacle Detection for In-Vehicle Sensor Fusion

Nikola Popovic (Faculty of Technical Sciences of Novi Sad, Serbia); Ognjen Djekic (Faculty Technical Sciences of Novi Sad, Serbia); Marko Kovacevic(Faculty of Technical Sciences, Serbia); Tomislav Maruna (RT-RK Computer Based Systems LLC, Serbia)
Advanced Driver Assistance System or ADAS is used in vehicles to make driving as safe as possible. For that reason ADAS uses sensors, lidars, radars and cameras to enable vehicles to perceive everything that surrounds them. So many different components require processing and analyzing large amounts of data. This paper proposes layered architecture for software which collects data from ultrasonic sensors, processes it and deploys to higher layers of the system.


Paper 6: A Solution to Software Scalability across ADAS Platforms

Ognjen Djekic (Faculty Technical Sciences of Novi Sad, Serbia); Veselin Gojak (Faculty of Tehnical Sciences, Serbia); Milos Ilic (Faculty of Technical Sciencis – University of Novi Sad, Serbia); Gordana Velikić (RTRK Computer Based Systems LLC, Serbia)
Recently Advanced Driver Assistance Systems (ADAS) use cases are growing in number and ubiquity with no common approach to software design and implementation to facilitate Advanced Driver Assistance Systems development. Thus we are presenting our scalable software model where we focus on data acquisition, processing units, and action intelligence. In this software module we present our Advanced Driver Assistance Systems Middleware design and communication with Advanced Driver Assistance Systems modules. The middleware design supports acquiring and processing data on different units with separation of concerns between the safety critical and non-critical data. This software model enables development of complex Advanced Driver Assistance Systems modules with higher efficiency based on standardized data types and data exchange between modules. The goal of this model is to help a researcher to reuse modules on different Advanced Driver Assistance Systems platforms.


Some BSc theses my students defended recently (in Serbian):

Kristina Omerović

Tijana Srejić

Miloš Ilić

Ivana Nikolić

Joakim Janjatović

Radivoje Ostojić

Nikola Popović