Search Results

Growing trends of connected and autonomous vehicles have pushed for increased resolutions of cameras to 8Mpix and displays to 4K/8K, leading to requirements for high-speed interfaces that support 10Gbps and beyond. Unlike data center or enterprise networks which normally operates under controlled indoor environments, in-vehicle networks are required to operate in harsh temperature and interference environments. Due to cost restrictions, the use of single pair wire is prevalent for in-vehicle networks. In general, as data transmission speed increase, signal spectrum spreads across greater frequency range. Since insertion loss of a channel increases in proportion to signal frequency, it becomes more difficult to secure SNR (signal-to-noise ratio) margins as bit rate increases. This makes it increasingly difficult for a device (e.g. ECUs, sensors, and displays) with high-speed communication interface to meet EMC (electromagnetic compatibility) criteria imposed by automotive OEMs.

Recent developments in the commercialization of mobility services have brought unprecedented connectivity to the automotive sector. While the adoption of connected features provides significant benefits to vehicle owners, adversaries may leverage zero-day attacks to target the expanded attack surface and make unauthorized access to sensitive data. Protecting new generations of automotive controllers against malicious intrusions requires solutions that do not depend on conventional countermeasures, which often fall short when pitted against sophisticated exploitation attempts. In this paper, we describe some of the latent risks in current automotive systems along with a well-engineered multi-layer defense strategy. Further, we introduce a novel and comprehensive attack and performance test framework which considers state-of-the-art memory corruption attacks, countermeasures and evaluation methods.

It is important that Advanced Driver Assistance Systems (ADAS) and Automated Driving Systems (AD) detect on-road objects, road vehicles and pedestrians. The typical detection devices mounted on ADAS and AD include a camera, a millimeter-wave radar and a Light Detection And Ranging (LiDAR). Since LiDAR can obtain accurate distance and fine spatial resolution due to its short wavelength, it is expected that small objects such as a tire can be detected. However, the conventional LiDAR is equipped with multiple light transmitters and light receivers such as avalanche photo diodes. This causes LiDAR system to be expensive and large in size. Aiming to reduce the cost and size of LiDAR, we employed Single-Photon Avalanche Diode (SPAD) which can be fabricated by CMOS process and easily arrayed. We also developed “Single Chip SPAD Array“ in which the two-dimensional array of SPAD and a signal processing block of range calculation were integrated into a single chip.

As automotive electronic systems aiming for a safe and secure automobile society continue to develop, the control specifications of the ECU are becoming increasingly complex. When attempting to validate control specifications that cooperatively control different control specifications, control specification developers must consider various validation viewpoints. They narrow down the validation viewpoints based on rules from past experience, describe timing charts, and validate the specifications. However, due to complicated specifications, empirical rules do not pass, and specification mismatches are often found after actual systems completion. On the other hand, a block diagram simulator is a tool to verify control specifications. Since these tools are aimed at modeling and verifying the system design, it is efficient to describe how to implement the system. However, first it is necessary to verify the consistency between the model and the specifications.

This paper presents an improvement in the accuracy of NOx sensors at high NOx concentration regions by optimizing the manufacturing process, sensor electrode materials and structure, in order to suppress the deterioration mechanism of sensor electrodes. Though NOx sensors generally consist of Pt/Au alloy based oxygen pump electrodes and Pt/Rh alloy based sensor electrodes, detailed experimental analysis of aged NOx sensors showed changes in the surface composition and morphology of the sensor electrode. The surface of the sensor electrode was covered with Au, which is not originally contained in the electrode, resulting in a diminished active site for NOx detection on the sensor electrode and a decrease in sensor output. Theoretical analysis using CAE with molecular dynamics supported that Au tends to be concentrated on the surface of the sensor electrode.

According to the advance of engine control development, demands for direct sensing of physical quantity have been growing. Regarding pressure sensing, key properties for direct sensing are robustness against high temperature and pressure, and response time in addition to accuracy. In this work, a pressure sensor module with these key properties was developed. First of all, a piezoelectric device was selected as a suitable sensing principle for the required properties because of its thermally stable piezoelectric effect and potential for simple installation structure. Regarding robustness against temperature, the sensor module was designed to form thermal isolation layer with outer housing which is optimized according to its application. Regarding robustness against pressure and response time, breakage of the piezoelectric element is the main technical issue.

Externally-connected Electronic Control Units (ECUs) contain millions of lines of code, which may contain security vulnerabilities. Hackers may exploit these vulnerabilities to gain code execution privileges, which affect public safety. Traditional Cybersecurity solutions fall short in meeting automotive ECU constraints such as zero false positives, intermittent connectivity, and low performance impact. A desirable solution would be deterministic, require minimum resources, and protect against known and unknown security threats. We integrated Autonomous Security on a BeagleBone Black (BBB) system to evaluate the feasibility of mitigating Cybersecurity risks against potential threats. We identified key metrics that should be measured, such as level of security, ease of integration and system performance impact. In this paper, we describe the integration and evaluation process and present its results.

As a method of maintaining thermal sensation and comfort inside a passenger compartment, not only a conventional HVAC system but also a combination of a HVAC system and other devices such as seat heaters, a steering wheel heater, ventilation seats are increasing. This research developed a method to evaluate thermal sensation of a human body when using these various thermal control devices. This method can evaluate the heat balance of the human body by calculating the amount of heat exchange between a human body and the external environment, and it takes into consideration the influence of heat exchange by heat conduction with seats or a steering wheel. The human thermal model is made by dividing a human body into various segments, and it is the model that considers heat transport by blood flow for each segment.

For the purpose of coping with the strengthening of NOx exhaust gas control and fuel consumption control, it is indispensable to improve the NOx purification capacity. In view of this, vehicle manufacturers are in the course of developing high performance SCR (Selective Catalytic Reduction) systems [1, 2]. For such SCR systems to be realized, high precision NOx sensors for carrying out urea injection quantity control and SCR degradation diagnosis are absolutely indispensable. Detection of NOx concentration by means of a NOx sensor is generally performed as follows: O2 is discharged by means of an O2 detection electrode; remaining NOx is decomposed by a NOx detection electrode; NOx concentration is then detected as electric current that flows when oxygen ions are conduct through solid electrolyte. In order to detect NOx of ppm-order, it is necessary to detect minute current of nA-order with high accuracy.

Compact, high efficiency and high reliability are required for an xEV motor generator. IPM rotors with neodymium magnets are widely applied for xEV motors to achieve these requirements. However, neodymium magnet material has a big impact on motor cost and there is supply chain risk due to increased usage of these rare earth materials for future automotive xEV’s. On the other hand, a wound-field rotor does not need magnets and can achieve equivalent performance to an IPM rotor. However, brushes are required in order to supply current to the winding coil of the rotor. This may cause insulation issues on xEV motors which utilize high voltage and high currents. Therefore, it is suggested to develop a system which supplies electric energy to the rotor field winding coil from the stator without brushes by applying a transformer between stator coil and rotor field winding. Specifically, add auxiliary magnetic poles between each field winding pole and wind sub-coils to these poles.

In gasoline direct injection (GDI) systems, various injection types are needed to reduce emissions and improve fuel consumption. This requires high-precision injection in the region in which the amount of injection is small. Achieving injection of a small amount of fuel using GDI solenoid injectors requires the use of the half-lift region. In this region, however, the variation in the injection amount tends to increase due to the variation in the lift behavior of the injectors, posing the problem of how to achieve high-precision injection. To reduce the variation, we analyzed the lift timing out of the injector current and voltage signal with the ECU in an attempt to adjust the amount of injection.

New 2A/F systems different from usual A/F-O2 systems are being developed to cope with strict regulation of exhaust gas. In the 2A/F systems, 2A/F sensors are equipped in front and rear of a three-way catalyst. The A/F-O2 systems are ideas which use a rear O2 to detect exhaust gas leaked from three-way catalyst early and feed back. On the other hand, the 2A/F systems are ideas which use a rear A/F sensor to detect nearly stoichiometric gas discharged from the three-way catalyst accurately, and to prevent leakage of exhaust gas from the three-way catalyst. Therefore, accurate detection of nearly stoichiometric gas by the rear A/F sensor is the most importrant for the 2A/F systems. In general, the A/F sensors can be classified into two types, so called, one-cell type and two-cell type. Because the one-cell type A/F sensors don’t have hysteresis, they have potential for higher accuracy.

Driving is an action that depends strongly on visual information. For displays in the cockpit, a combination of “ease of viewing” to inform the driver of danger early and “annoyance reduction” to avoid drops in the driver’s perception is needed. In this study, we tried to capture “ease of viewing” and “annoyance” in one fixed-quantity indicator. We took up a Camera Monitor System (CMS) as the subject and analyzed the effect that annoyance with the display used in CMSs has on driving behavior. Based on our analysis, we hypothesize that evaluating carelessness in viewing behavior is related evaluating to annoyance. Next, we chose a Detection Response Task (DRT) technique as a method to evaluate driving behavior influenced by this annoyance.

Recent electric vehicles use Li-ion batteries to power the main electric motor. To maintain the safety of the main electric motor battery using Li-ion cells, it is necessary to monitor the voltage of each cell. DENSO has developed a battery Electronic Control Unit (ECU) that contributes greatly to the reduction of the cost and the improvement of the reliability of the system. Each manufacturer has been developing a dedicated IC for monitoring the voltages of each cell of a battery. However, since the number of cells that can be monitored is limited, more than one IC is required to measure the voltages of a large number of cells. The increase in the number of ICs and the amount of insulator leads to the rise in system cost. DENSO has developed a dedicated IC that uses a proprietary high-breakdown voltage process, and which enables monitoring up to 24 cells with a single IC chip.

In-vehicle network communication is evolving faster speeds and higher performance capabilities, connecting the information possessed by ECU and sensors with the in-vehicle electronic systems which are continuing to develop. With the evolution of the complicated networks, it is becoming difficult to develop them without many verification of actual machine. On the other hand, as for the verification means required at the logic level or physical level for a network verification through ECU design, virtual verification in the whole vehicle is difficult due to speed increases and the sheer size of the system. Therefore, it is only applicable for systems which are limited to a domain or an area, and flexible and timely utilization would be difficult due to the changes in specifications.

In the last decade, radar-based Advanced Driver Assistance Systems (ADAS) have improved safety of transportation. Today, the standardization of ADAS established by New Car Assessment Program (NCAP) is expected to expand its market globally. One of the key technologies of ADAS is the rear-side monitoring system such as Blind Spot Warning (BSW) and Closing Vehicle Warning (CVW). It is required to expand its detection range so that it can monitor not only nearside targets for BSW, but farther targets for CVW. These applications can be achieved using two radar sensors installed at rear-side corner of the vehicle. However, the expanded detection range causes undesirable target detections and decreases target recognition performance. In this paper, a novel solution to improve the performance using DCMP(Directional-Constrained Minimization of Power)-based Beamspace technology using Two-frequency continuous wave (2FCW also known as FSK) is introduced.

A Rear Cross Traffic Auto Brake (RCTAB) system has been developed that uses radar sensors to detect vehicles approaching from the right or left at the rear of the driver’s vehicle, and then performs braking control if the system judges that a collision may occur. This system predicts the intersecting course of approaching vehicles and uses the calculated time-to-collision (TTC) to control the timing of automatic braking with the aim of helping prevent unnecessary operation while ensuring system performance.

The electric power steering (EPS) is increasing its number since there are many advantages compared to hydraulic power steering. The EPS saves fuel and eliminates hydraulic fluid. Also, it is more suitable to the cooperation control with the other vehicle components. The EPS is now expanding to the heavier vehicle with the advance in the power electronics. In order to meet customer's needs, such as down-sizing, lower failure rate and lower price, we have developed the new motor control unit (MCU) for the EPS. The motor and the electric control unit (ECU) were integrated for the better installation. We adopted new technologies of redundant 2-drive design for more safe EPS. “2-drive Motor Control technology” which consists of dual winding, two torque sensors and two inverter drive units. In our developed MCU, even if there is a failure in one of the drive unit, the assistance of the EPS can be maintained with the other drive unit.

Modern vehicles utilize various functionalities that require security solutions such as secure in-vehicle communication and ECU authentication. Cryptographic keys are the basis for such security solutions. We propose two approaches for secure and efficient invehicle key management. In both approaches, an ECU acting as a Key Master in the vehicle is required. The first approach is based on SHE. The Key Master generates and distributes new keys to all ECU based on the SHE key update protocol. The second approach performs key establishment based on key derivation. The Key Master sends a trigger in form of a counter and all ECUs derive new keys based on the received counter value and pre-shared keys. It is thus possible to handle in-vehicle key management without the need for an OEM backend to manage all keys. This reduces cost and complexity of the solution.

This paper presents detection technology for a driver monitoring system using JINS MEME, an eyewear-type wearable device. Serious accidents caused by human error such as dozing while driving or inattentive driving have been increasing recently in Japan. JINS MEME is expected to contribute to reducing the number of traffic deaths by constantly monitoring the driver with an ocular potential sensor. This paper also explains how a driver’s drowsiness level can be estimated from information on their blink rate, which can be calculated from the ocular potential.