International Conference on Data Mining and Artificial Intelligence Methods for Automotive Systems

Conference and Training Center ,,CIS’’ in Szczyrk, Poland

May 23-24, 2016

Abstracts

Information Models – The Foundation for the IIoT!?
Daniel Grossmann
THI

Simulation Based Verification of Active Safety Algorithms for High-Risk Maneuver Applications
Rafał Burza, Dariusz Cieślar, Krzysztof Kogut
AGH

An unnecessary triggering of automotive Active Safety (AS) functionalities is always noticeable to the driver, because they are designed to attract driver’s attention or to intervene by sending commands to actuators. Suppression of false positive triggering is therefore of high priority as they can cause considerable disruption to driving comfort. This work considers the development of production stage algorithms operating based on locally collected information about vehicle surroundings. Such information is extracted from AS sensors by means of perception algorithms. The resulting description of the dynamically changing operating conditions is subsequently used by particular AS functions, which take appropriate actions. The development process of underlying algorithms shall include testing phase during every project stage in order to limit possibility of inducing an error into the system. Experimental verification of such functionalities is always a necessary, ultimate test, but in case of collision mitigation concepts experiments are inherently costly, potentially dangerous and often non-exhaustive. Industrial state of the art approaches involve simulation of sensors, e.g. radars, lidars or vision based sensors, and processing their outputs by production code prototypes. The use of realistic virtual scenes containing sensor models can greatly improve the speed of development and allow more extensive testing. The article provides fundamentals of some most common sensor models in Active Safety domain, the explanation of developed test cases and results of simulations. A simplified Rear Collision Detection functionality is considered as an illustrative example.

Computer for car sport
Krzysztof Tokarz, Piotr Czekalski, Roderyk Raszka
SUT

Article presents project of the computer that can be used in amateur car races. Main functions of this computer are to count laps, measure lap time and log race parameters as overloads, pitch and roll of the car. Computer uses GPS module, accelerometer and magnetometer sensors.

Development of remote control device and transmission protocol
Krzysztof Tokarz, Michał Mączka
SUT

Main goal of this research is to increase communication range of AutoUniMo device based on WIFI. WIFI provides maximum remote control at distance about 20 m. The 433 MHz RF based protocol enables remote control on distance 200 meters or higher. This survey compares most popular RF modules available in respect of range, and also checks influence of obstackles like houses or trees on remote control. All researches were made on example wireless control of caterpillar based robot with manipulator. Whole work focuses also at hardware procotols and functions implemented on this RF modules.

Technical solutions for modern road traffic
Sonja Waechter, Hueseyin Erdogan, Adam Ziębiński, Rafał Cupek
Continetal, SUT

Traffic has become more complex in recent years and therefore the expectations that are placed on automobiles have also risen sharply. Increased mobility, unclear traffic and traffic sign signage, time pressure and the desire for continuous availability, make it increasingly difficult for humans to cope with the traffic situation. Support for drivers and the protection of the occupants of vehicles and other persons involved in road traffic have become essential. Rapid technical developments and innovative advances in recent years have enabled the development of plenty of control units that are based on different working principles such as radar, lidar or camera techniques. Some systems only warn the drivers via a visual, audible or haptical signal of a danger. Other systems are used to actively engage in the control of a vehicle in emergency situations. Some application functions of Advanced Driver Assistance Systems (ADAS) are, for example, Blind Spot Detection, Traffic Sign Recognition, Emergency Brake Assist and several others. All sensors have their advantages and disadvantages, which means that the most efficient applications use a combination of different sensor types. These sensors are no longer a luxury item for upscale vehicles but are standard equipment in low-cost cars. Although technical development is already quite mature, there are still many development opportunities for improving road safety. The further development of current applications and the creation of new applications that are based on sensor fusion are essential for the future. Only through the stronger networking of sensors and the entire view of road traffic, including communication between the various road users, can the pursued objective of autonomous driving be realised.

Model of a production stand used for digital factory purposes
Marek Drewniak
AIUT

Realisation of production lines and stands in industrial facilities is related to interconnection of many realisation fields: from physical construction, through control and measurement systems, to offline analysis and documentations. Very often the mechanisms and technologies that were prepared once are used as a knowledge base in creation of new production structures. The transfer of knowledge is usually realised as an exchanging of good practice or as reuse of formerly prepared technology for selected component. In order to unify the description of a whole production stand along with all relations between its fields there was a new concept method defined as a digital factory. This approach describes selected part of a production system (the whole stand or its fragment, the production line or the facility) as a model that combines realisation components with dependencies that exist between them: common usage of components of a station in individual fields, their roles and history of changes. The approach allows to gather experiences gained during launching procedures and use them in construction process of systems that are somehow similar. Additionally, it provides quick and direct access to resources related to the field in which the changes are considered (e.g. during technological modifications or in maintenance actions).
In the presentation a typical structure of a production stand is presented, along with fields that are commonly used in production-measurement technology. Then, a life cycle of a stand is described, which consists of realisation stages: from work assumptions to maintenance. The introduction to usage of different types of resources related to each field and each stage is presented. Subsequently, a concept of digital factory model is described. The solution considers realisation fields mentioned previously and explains dependencies that may occur between components of a stand (how components of a system can intersect in vary fields). Finally, two use cases of a model based on a real examples are explained.

Digital data in the product life cycle
Michael Stempfle, Markus Bregulla
THI

The constant increase in the price of energy, more restrictive emission norms and rapidly growing recycling charges cause new challenges for production sites. Especially in Europe the energy prices have a high impact to the factory costs. One missed chance is, that production companies don’t use longer periods of no manufacturing to switch to a high energy saving state. These longer periods of non-production time include weekends and production shutdowns during the holidays. This energy saving potential is often unexploited. Hence the base load can reach up to 50% of the energy required during production phases. The aim is to decrease the non-value energy consumption to a minimum. On the other side it is important to guarantee the production availability. After switching to such an energy saving state it is essential to ensure a smoothly restart of the production.
The following focus on automated automotive production lines and approaches to reduce non-value energy consumption. It is also absolutely essential to monitor the energy consumption on a regular basis.

Cloud based services for information flow in industrial applications
Sebastian Schrittenloher, Markus Bregulla
THI

Target of the project is, to create an intelligent platform where information can be stored. Knowledge, that is once gained, should be stored in this platform. Synergies between finished and new projects should be reached. The platform should be created by using cloud computing.
In special it is for data of energy consumption and how reduces can be achieved. Once applied measures to save energy, should work also a second time on another place, or could be improved. For the storing system cloud solutions should be used, because much space for a big volume of data will be needed and to get account on every place. The data in the cloud must be consistent. Different programmes will be used for simulation (e.g. CATIA, WinMOD). They will use the same data and if the data in one programme will be changed or actualised, the other programmes, that use this data, must be changed to the same value. Therefore software solutions will be necessary. Actual connections in clouds are not very far developed. A solution for this problem will have to be found. Actual cloud solutions from different suppliers are already in check. Another point is the safety of the data in the cloud. To get it save from abuse, algorithms are necessary that can ensure that the cloud is safe. For this problem are actual no practicable solutions available. Possibilities to ensure the safety will have to be searched. To improve the data processing inside the cloud, other opportunities of clouds will be checked.

LiDAR technology in general purpose applications
Adam Ziębiński, Rafał Cupek, Michał Kruk, Marek Drewniak
SUT, AIUT

Technologies that support safety on the road and in the car traffic are currently being developed very rapidly. This causes the birth of the whole branches of solutions for specific applications. The systems that support driver are popularly called Advanced Driver Assistance Systems (ADAS) and are constructed based on measurement devices that work on CAN bus. One of such device is a lidar. The lidar is a module built of an emitter that sends laser beam and a receiver with a telescope, which is used for measurement of distance and calculation of approach velocity of an object. There are many solutions that use this technology, not only in the automotive systems. For the case of distance measurement to the obstacles on road there are short range lidars used mostly, which use narrow range of wave length near infrared. The lidar, along with additional numerous systems and onboard devices, is used for support during slow car movement, e.g. during remaining in traffic jams for keeping the safe distance to the vehicle to the front or in automatic braking systems.
In the paper authors briefly present overview of technology and LiDAR devices and introduce CAN technology which is used in the automotive networks. Then, an example of operation and startup conditions of lidar used in ADAS systems are presented. Due to transmission of specifically defined data that come from the external devices, a usage of lidar in continuous measurement mode is presented. During this mode, the lidar provides information about distance measured on three independent channels, velocity of an approaching object and the amount of light reflected from the obstacle that travels back to the receiver. Those allow to expand the use of the lidar outside the field of automotive applications. In experiment section, there is a laboratory stand described on which the tests were performed. Additionally, capabilities of a device were shown along with short analysis of their use.

Simulation in CANoe on the basis Lidar
Marek Drewniak, Hueseyin Erdogan, Adam Ziębiński, Rafał Cupek, Michał Kruk
AIUT, Continetal, SUT

With the growing number of control units and devices working on electronic control unit (ECU) networks the need of advanced software which allows analysis of ECU increase. Use of such a software not only allows the testing of devices itself, but also simulate their environment which can also be really wide. The presented software from Vector company called CANoe is comprehensive software tool which allows development, testing and analysis of whole ECU. Models created in CANoe can identify and correct problems at an early stage. They can be used as the basis for analyzing, testing and integrating bus systems. There are both graphically oriented and text-based evaluations. The simple simulation on the basis Lidar shows main functionalities like CAN units simulation, events programming or graphic interpretation of data, which can be made using CANoe.

GPS and ultrasonic distance sensors for Autonomous Mobile Platform
Wojciech Czernek, Wojciech Margas, Roman Wyżgolik, Sebastian Budzan, Adam Ziębiński, Rafał Cupek
SUT

Thanks to today’s technology the dream of fully automated, self-driving car is just around the corner. A number of international giants like Microsoft or Tesla are working tirelessly to make this dream come true.
Following their footsteps the students of Silesian University of Technology started the AutoUniMo project.
The goal of the project was to create an autonomous mobile platform based on Raspberry Pi which would be able to scan its surroundings and navigate autonomously. To achieve that goal the platform was equipped with a vast number of various sensors.
In this paper the connection of GPS sensors and ultrasonic range meters is described together with performed tests and comparisons between different sensors of each type.
The GPS sensors chosen for the project were: PmodGPS from Digilent Inc. and GY-GPS6MV2. First one is based on MediaTek GPS MT3329 with internal antenna and a possibility to connect an external one. The latter is based on Ublox NEO-6M-0-001 GPSmodule with external antenna. Both GPS modules use UART interface for communication with the Raspberry Pi. The software for both units was developed and the measurements of sensors accuracy were performed.
For the purpose of the project also two models of ultrasonic distance sensors were chosen. First of them was HC – SR04. It is a cheap ultrasonic distance sensor with a simple mode of operation and measuring range up to 4 meters. The second sensor was URM37 V3.2. The sensor is based on industrial level AVR processor and has a number of operating modes and temperature correction which is uncommon among similar devices. The software for both sensors was developed and the accuracy of their measurements was tested.
Thanks to the sensors described in the paper the mobile platform is able to navigate on longer distances using satellites as well as in closed narrow spaces.

APPLICATION OF ISA95 Data models IN Manufacturing Execution Systems FOR LEAN Production
Udo Wozar, Huseyin Erdogan, Rafal Cupek, Szymon Ziemek
Continetal, SUT

Manufacturing Execution Systems (MES) are service-oriented interfaces that join the word of business transactions with the world of production systems. Nowadays IT systems have to provide very detailed information that is related to an underlying production process and also to actual product. There are a few emerging business models that require accurate and timely production data. This document examines two approaches to database architecture that can be used in Manufacturing Execution Systems (MES). It focuses on the support of the LEAN business model. The main research goal is to support the flexible access to production data, but the efficiency of database is also very important factor. Authors compare the classical relational database model with the object-oriented one. Considered use cases include the Oracle DB and Oracle Objects applications for MES. Presented object oriented approach follows the ISA95 standard. The practical use cases are based on the production of electronic devices carried out by the company Continental Ingolstadt. Although object oriented databases are not well accepted by the industry due to their low efficiency, the authors show that in the case of LEAN production, the database system based object-oriented models can be far more convenient than a classical relational database. The main benefits are more flexible data model and highly adjustable MES that can follow changes in the underlying production system. By the case of LEAN manufacturing, authors show that the flexible object oriented database is more efficient solution comparing to the relational database. Moreover such an approach can help to avoid well known big data problems that are common in classical MES.

An EDDL (Electronic Device Description Language) based approach for communication with DBMS and File System in an Industrial Automation Scenario
Suprateek Banerjee
THI

In an industrial automation scenario, there is often a need to access a DBMS (Data base management system) and/or a local or shared file system for various purposes. EDDL (Electronic Device Description Language) provides a standardized way to access and exchange information with the devices that make up the automation system. This paper discusses an approach to use the EDDL approach to describe databases and files, which is to say that in this approach the databases and files both have been modelled as ‘devices’ and can be communicated with like any other device with its own device description. For this to work EDDL based device descriptions were formulated for databases as well as files and the communication with the same was established with the help of an OPC UA Server, containing nodes corresponding to database entries and files (of a local file system), in its address space. The OPC UA Server used the EDDL based descriptions of the file system and databases to gain connection and access related information about the same, thus establishing the communication. The results of this approach have also been discussed in the following sections.

The OPC UA information model for the support of the energy efficient production
Suprateek Banerjee
THI

In today’s manufacturing and production scenario, the term ‘energy efficiency’ is gaining much importance by the day. There is a need to monitor the energy efficiency of the system on a regular basis in order to maintain the energy standards of today’s industry. In addition to this, smart monitoring and diagnostics of the energy consumption of the system components also provide a valuable insight on the different energy consuming processes or cycles which constitute the overall system as a whole. When the energy consumption of the whole production system could be narrowed down or split into several energy consuming modules, it would be also possible to point out a particular module which might be showing erratic energy consumption patterns, thereby signifying a possible fault in the components involved in that particular module. Therefore there is a need to first identify and model such energy profiles of the whole system and identify the consumption pattern with the help of historical data about the same. This presentation outlines such an approach of modelling the energy efficiency of the system with the help of OPC UA, and since OPC UA inherently provides us with features like Historical Access for its data variables, this approach would take advantage of the same to collect the data in an external database in order to be analyzed with the help of Data mining techniques to find the energy consumption patterns within the collected data.

Power management and sensors handling on the Autonomous Mobile Platform in AutoUniMo project
Paweł Rybka, Łukasz Szczepański, Krzysztof Wosik, Roman Wyżgolik, Sebastian Budzan, Adam Ziębiński, Rafał Cupek
SUT

Nowadays there is a tendency of automating everyday life subjects. For several recent years there exists a conception of fully automated driverless vehicle so projects and experiments concerning the topic are run worldwide. Our project of autonomous car is rooted on the Raspberry Pi – Linux based embedded computer which gathers data from sensors, performs all calculations and runs operations on external modules.
For the platform to fully operate we had to connect numerous modules. The ones our team was responsible of were accelerometer (ADXL345), magnetometer (GY-271), gyroscope (PModGyro) and a wheel encoder (dedicated to DFRobot). Gathering and processing information from the modules allow the platform to calculate its position, heading direction, velocity, acceleration, angles, distance travelled and many more. This data is crucial for the vehicle to be fully self-operational as this intelligence is then converted into information which Raspberry uses to decide of its further decision making. Sensor system build on the Raspberry and set of mentioned sensors can measure, gather and send collected data over TCP protocol. Thanks to that platform gets data that defines its state and position in space. This data next can be used to perform more complex operation (for example calculating turning angles or regulating engine power according to changeable inclination of the road). Moreover, to extend the operational time of batteries (the platform is powered with Lithium-Polymer packs) there was created a printed circuit board - Raspberry HAT (Hardware Attached on Top) which enables the vehicle to control the power of particular devices. Even if a certain module or engine is unused it still draws current from the batteries, but with the programmable “key” Raspberry can decide to turn off futile components, and when is needed – turn them back on.

Passing data between processes in linux system, Wojciech Michna
Adam Ziębiński
SUT

Presentation about passing bulks of information’s in system is built from many smaller applications. Solutions for the exchange of information between applications, which some of these applications have to run in real time scheduler or be run as daemon. Presentation method of passing data from kernel-space to user-space as well as.

Presentation of OPC – UA implementation in Node JS
Wojciech Michna, Rafał Cupek
SUT

Introduction to Node Js implementation of OPC-UA Stack. Presentation about preparing and starting OPC-UA server and client in Node JS environment. During this presentation you will learn how to install packets, implement server and client side for OPC-UA stack, begin read write operations and establish subscription. Demonstration of portability of this solution by showing communication between two micro computers with different versions of NodeJS engine.

Data acquisition system for Autonomous Mobile Platform, based on STM32 Microcontroller
Adam Piech, Adam Ziębiński
SUT

The data acquisition system for Autonomous Mobile Platform is currently based on few sensors connected directly to the Raspberry Pi computer. As the number of different sensors grows, the limit of devices connected has been reached. To solve this issue, an extra node, that will collect this measurements in real-time is needed. Such solution allows to preprocess data and unitize it’s readout. Planned device, based on STM32 microcontroller, opens new possibilities for mobile platform development. Allowing connection for more sensors, same unit will be controlling DC motors and serving as a CAN interface for the Raspberry Pi. A proper software architecture must be assured, due to possible changes in hardware in the future. Some more advanced functionalities, like maintaining travel direction might be implemented, but the main goal for the device is to expand Raspberry Pi functionality, by performing periodic readouts, such as measuring distance in various directions, wheels rotation speed and collecting data from other components (e.g. power consumption meter, GPS), combining this low-level functionalities all together.

Drive control
Sebastian Kłak, Adam Ziębiński
SUT

hoosing right drive for mobile platform is not so simple task as it looks like. There are various kinds of drive types in view of its application and capabilities. Each solution has its advantages and disadvantages such as good stability, low energy consumption, good torque or complex build. With all these information we can pick right drive for our requirements. For simple prototype platform there is no need of its power and low energy consumption, so cheap already assembled universal platform is good enough, but for final version there are requirements and platform has to be built by ourselves. Tough Robotwerks Forbot platform seems to meet our assumptions.