A collaboration between AVL and Rohde & Schwarz, two providers of measuring and automotive testing systems, now permits the reproduction of realistic GNSS reception conditions for testbed vehicle testing. As a result, users can reliably test all aspects of GNSS-based vehicle positioning — a core functionality of autonomous vehicles.
AVL DRIVINGCUBE enables the reproducible testing of driver assistance systems and driving features for self-driving vehicles using a real vehicle within a virtual environment in a variety of different traffic situations. For that purpose, test drives are performed with a real, ready-to-drive vehicle on a chassis dynamometer or powertrain testbed.
With the help of realistic virtual driving scenarios, it is possible to test peripheral sensors, control systems and actuators inside the vehicle in a fully reproducible and reliable manner. Automated vehicle functions are thus sufficiently validated during development and even before testing on the proving ground.
The range of environment simulations carried out with AVL DRIVINGCUBE can now be extended to include GNSS signals, bringing simulation closer to reality than ever before. The vehicle’s GNSS receiver is stimulated realistically using GNSS signals generated on the testbed.
This way, technical engineers can identify exactly how sensors, automated driving features and other actuators respond inside the vehicle. The now possible GNSS-based vehicle positioning feature is a core functionality of automated driving, and the approach ensures that it is reliably tested.
The SMBV100B GNSS simulator. (Photo: Rohde & Schwarz)
For generating GNSS signals, Rohde & Schwarz GNSS simulators are used (R&S SMBV100B or R&S SMW200A), which allow the generation of signals for all of the available satellite navigation systems (GPS, Glonass, Galileo, BeiDou, QZSS, SBAS) across all frequency bandwidths (L1, L2, L5). This also makes them suitable for testing multi-frequency receivers, which are playing an increasingly important role in automated driving.
“In Rohde & Schwarz, we now have a strong and reliable partner for GNSS stimulation. By generating consistent GNSS signals in connection with environment simulation, AVL DRIVINGCUBE now provides a test system that allows users to validate GNSS-based driver assistance systems and autonomous driving features,” explains Dr.-Ing. Tobias Düser, Head of Advanced Solution Lab at AVL Deutschland GmbH.
Christoph Pointner, Head of Signal Generators at Rohde & Schwarz, adds: “We are very pleased to bring our expertise in the field of signal generation to this collaboration with AVL and contribute to such an important innovation and trendsetting solution for testing automatized driving features.”
The additional GNSS stimulation makes testbed testing not only more realistic, it is above all a further step in moving testing from the road to the rig. This leads to a much sharper reduction of test drives than was the case previously and major savings in the kilometers driven.
Rohde & Schwarz GNSS stimulators form a flexible, modular system that can be adapted to your requirements and is easily integrated in the AVL DRIVINGCUBE environment. The stimulator is controlled automatically from the simulation platform. GNSS extensions for AVL DRIVINGCUBE are available with immediate effect.
AVL DRIVINGCUBE enables the reproducible testing of driver assistance systems for self-driving vehicles. (Photo: AVL)
GNSS chip manufacturers and positioning systems developers are working on bespoke devices for autonomous driving. This month, we look at a development with embedded functional safety.
By Fabio Pisoni, Domenico Di Grazia, Giuseppe Avellone, Luis Serrano, Brett Kruger, Laura Norman and Natasha Wong Ken
INNOVATION INSIGHTS by Richard Langley
I DRIVE A 10-YEAR OLD KIA SPORTAGE. It is still quite roadworthy despite having to contend with New Brunswick winters. However, it lacks some of the safety features that are present in newer cars. There is no back-up camera, no forward-collision warning, no automatic emergency braking, and no blind-spot warning, for example. These are just some of the safety systems that come as standard or optional on most new cars these days. Still, the driver is responsible for the safety and operation of the car at all times. True, help might be provided for parallel parking and cruise control, but that’s about it for automated operation with most vehicles.
But things are changing and changing fast. Real automation is coming to automobiles. Already partial automation is available in some high-end vehicles that can take over steering, braking and acceleration in certain circumstances. The driver is still responsible for other aspects of the vehicle’s operation including paying attention to road conditions. Soon, we will have conditional automation where the car can drive itself but the driver must stay alert and be prepared to take over immediately at any time. Next will come high automation where a computer fully drives the car at certain times on certain routes such as a highway. Multiple systems, including back-up systems, will maintain a required safety level and the car will determine if it is safe to operate autonomously. If not, it could pull over to the side of the road and shut down. And finally, we may have full automation of cars. They will be able to drive on any road under virtually any conditions and won’t need any controls such as steering wheels or accelerator or brake pedals.
Augmented GNSS guidance will play a major role in the automation of vehicles. As with any navigation or guidance system, there are four important requirements: accuracy, availability, continuity and integrity. Perhaps the most obvious requirement, accuracy describes how well a measured value agrees with a reference value, typically the true value. How well a system accounts for various errors or biases determines the accuracy of corrected measurements and, ultimately, the accuracy of a derived position. A navigation system’s availability refers to its ability to provide the required function and performance within the specified coverage area at the start of an intended operation. In many cases, system availability implies signal availability. Environmental factors such as signal attenuation or blockage or the presence of interfering signals might affect availability. Ideally, any navigation system should be continuously available to users. But, because of scheduled maintenance or unpredictable outages, a particular system may be unavailable at a certain time. Continuity, accordingly, is the ability of a navigation system to function without interruption during an intended period of operation.
While accuracy, availability and continuity of a guidance system are all important, it is the integrity or trustworthiness of the system that is paramount. It is why the automotive industry has already developed integrity standards for the automation of vehicles. And it is why GNSS chip manufacturers and positioning systems developers are working on bespoke devices for autonomous driving, whatever the level of automation. In the Innovation column this time around, we’ll learn about one such development — one with embedded functional safety.
Autonomous driving applications are raising the requirements for onboard GNSS receivers to new highs. Position accuracy, protection levels, high availability, robustness of operation and integrity are the priorities shaping a new class of automotive components and architectures. Autonomous driving deals with life-critical issues: the expectation of reliability and safety for this new generation of receivers, as well as for other sensors and systems, is very high.
The International Organization for Standardization (known by the language-independent short form ISO) has issued documents codifying functional safety (FuSa) for automotive applications: ISO 26262: part 1 to part 11. ISO 26262 complements the well-known automotive reliability standard published by the Automotive Electronics Council, AEC-Q100. With respect to FuSa, a system can be defined as functionally safe if it always operates correctly and predictably. More importantly, in the event of failures, the system must remain safe for people. Lastly, as security is becoming paramount, a new standard for cybersecurity in automotive applications — ISO/SAE 21434 — is in development by ISO and SAE International (initially called the Society of Automotive Engineers) that will require a GNSS receiver to be robust against jamming, spoofing and meaconing attacks.
The Automotive Safety Integrity Level (ASIL) is a key part of ISO 26262 compliance, and the standard specifically identifies the minimum testing requirements depending on the ASIL of the component. The ASIL of a component or system depends on the ASIL of the target application. The ASIL is determined at the beginning of a development process. It varies from ASIL-A to ASIL-D, where A is for less critical applications and D for the most critical ones such as steering and breaking systems. ASIL-rated lane-level positioning performance can be demonstrated today by combining an ASIL-B software positioning engine and TerraStar-X correction technology from Hexagon Positioning Intelligence with GNSS measurements from an ASIL-B-rated GNSS chipset.
To conjugate performance requirements with the demand of embedded functional safety, STMicroelectronics has developed TeseoAPP (STA9100), a next-generation GNSS component, designed to meet an ASIL-B level of safety. TeseoAPP is a multi-band GNSS measurement engine. It outputs all the observables, navigation and integrity data required by a safety-critical precise positioning algorithm, located on a host processor. TeseoAPP also computes a local L1 code-based standard position, velocity and time (PVT) solution (SPS) for monitoring and integrity purposes. Also part of the baseline features are autonomous satellite acquisition (cold start condition), real-time assistance, data decoding and storage on external non-volatile memory (NVM), accurate timing and pulse-per-second generation under vehicle dynamics.
RECEIVER ARCHITECTURE
The target architecture for a safety-critical platform is sketched in FIGURE 1, where a host microprocessor is in charge of collecting GNSS observables and sensor data from the TeseoAPP. The latter includes on the same chip die a first configurable RF chain for the L1 signal ensemble and the baseband part for processing all the signals in the served bands, while the second chip is an RF front end (code-named STA5635), configurable for receiving the other served bands (such as GPS L2 or L5, Galileo E5a or E5b or E6, and so forth). The two chips are clearly visible in the photograph of a TeseoAPP evaluation module of FIGURE 2.
FIGURE 1. Block diagram of the TeseoAPP platform for safety-critical applications, featuring surface-acoustic-wave (SAW) filters, a temperature-compensated crystal oscillator (TXCO), non-volatile memory (NVM) and both internal and external STA5635 tuners. (See text for other initialisms used.) Diagram: Authors)FIGURE 2 The TeseoAPP Evaluation Module, including the STA9100 (TeseoAPP) and STA5635 (external tuner). Photo: Authors
The selected frequency plan and constellation configuration depend on the specific autonomous driving scenario and the target geographic area. The TeseoAPP supports a mix of frequencies and signals as shown in TABLE 1. The chipset baseband unit can track up to 80 channels. A tracking snapshot from a rooftop antenna (located at the ST office in Naples, Italy) is illustrated in FIGURE 3.
Both the TeseoAPP and the STA5635 have been designed for ASIL-B following the concept of “safety element out of context” (SEooC) described in ISO standard ISO 26262:2012. In this context, assumptions have been made for the application (such as on the mission profile), identifying the related safety goals from which functional and technical safety requirements have been derived.
TABLE 1. The TeseoAPP (STA5635) supported frequency plans and scenarios.FIGURE 3. Screenshot of the L1-L5 TeseoAPP configuration, from the ST Teseo-Suite tool (using the Naples rooftop antenna). Image: Authors
Following the guidelines identified in the ISO 26262 flow for safety-relevant product development, several safety mechanisms have been identified at the hardware, firmware and system/boot level. The microcontroller unit (MCU) supports dual-core operation in a lock-step configuration to verify processor output errors together with a memory built-in self-test (executed at startup) and error correction code on a safety-related embedded random access memory. Other hardware redundancies have been introduced in safety relevant parts such as triple-voted registers for critical configuration parameters. For the real-time operating system (RTOS), an ASIL-D-level product — the highest level — was selected.Functional safety analysis of the GNSS sub-system has produced a dedicated technical safety concept, including aspects such as tuner operation, interference and jamming mitigation, signals and observables quality management (QM), reliable host communication (using generic end-to-end or E2E protocols for data integrity and resilient flow control), and reliable system software. A simplified overview of all these safety mechanisms is outlined in FIGURE 4, where the orange-colored blocks are specific for the GNSS sub-system.
FIGURE 4. Overview of the TeseoAPP safety mechanisms. (See text for acronyms and initialisms used.) Diagram: Authors
Safety Mechanisms. The technical safety concept of the GNSS sub-system is implemented by a security, integrity and safety (SIS) monitoring layer (see FIGURE 5). The SIS collects information and metrics from other receiver blocks embedded in the RF/baseband hardware and from different components of the GNSS firmware stack. The SIS internally computes integrity risk estimates, which are delivered to a central intelligence monitor (CIM) capable of switching the receiver into a safe state, within a fault-tolerant time interval, when the overall receiver integrity appears compromised. In its simplest form, the CIM can be represented by a weighted sum of integrity risk inputs, followed by some activation function. During this process, a first layer of logic (CIM-L1) combines a subset of signal quality metrics to decide a priori which observables shall be passed to the host or discarded (not delivered).
FIGURE 5 Safety information flow through the TeseoAPP security, integrity and safety layer. (IP = intellectual property; other short forms in text.) Diagram: Authors
The collected signal metrics include quality estimators (based on multi-correlation techniques for example) or classic linear combinations of observables (such as dual-frequency carrier-phase differences or code-minus-carrier). Receiver metrics, on the other hand, have a more global scope and include estimators for inter-frequency biases, system-time cross-checks among constellations, and so on. The fault collection and control unit (FCCU) conveys hardware status flags to the SIS. Typically, an FCCU exception indicates some critical hardware failure and takes a priority path when switching the safe state. For example, a fault in the MCU lock-step monitor will trigger an immediate firmware action, mediated by the FCCU.
POSITIONING PERFORMANCE
To demonstrate the performance that can be achieved using the ST TeseoAPP chipset, Hexagon Positioning Intelligence (PI) has combined measurements from the TeseoAPP with an automotive-grade antenna and Terrastar-X correction technology, and processed the data using Hexagon PI’s software positioning engine. Even with a modern receiver supporting dual-frequency, multi-constellation measurements, such as the TeseoAPP, corrections are necessary to deliver decimeter-level performance and safety information required by an autonomous vehicle.
In clear-sky environments, lane-level positioning accuracy is achieved, enabling GNSS as a key input to autonomous systems. FIGURE 6 shows the horizontal error performance of the combined ST+PI solution in the form of an error time series and an error cumulative distribution function (CDF). The error performance expected from today’s single frequency automotive-grade GNSS without corrections and processing is also shown for comparison.
FIGURE 6. Horizontal error time series and cumulative distribution function (CDF) of the TeseoAPP alone and of the TeseoAPP with Hexagon PI software positioning engine (SWPE) in an open-sky environment. (Image: Authors)
For guidance systems in autonomous applications, the GNSS position must be accompanied by safety information and integrity guarantees. The concept of protection levels (PLs) has been introduced to provide this. A horizontal protection level defines a circle or ellipse around the reported GNSS position, which will have some error, within which the actual position is guaranteed to fall. The Hexagon PI software positioning engine is ASIL-B rated, so its position and PL outputs are available for use in safety-related autonomous applications. The autonomous system using the GNSS position is assured that its actual position is within the protection level ellipse. To output ASIL-B-rated positions accompanied by PLs, ASIL-rated GNSS measurement inputs are required.
Using the inputs and techniques described above, the Hexagon PI software positioning engine calculates PLs for every GNSS position output. The Hexagon PI data from Figure 6 is shown again in FIGURE 7 with accompanying PL information. In this case, a PL with integrity risk of 10-7 is shown, meaning that the actual position error is expected to exceed the reported PL at a rate less than 10-7 per hour.
FIGURE 7. Horizontal error and protection level (PL) including cumulative distribution functions (CDFs) of the Hexagon PI software positioning engine (SWPE) in an open-sky environment. (Image: Authors)
The PLs shown in Figure 7 are typically much greater than the position error. This is because the protection level calculation must account for a large number of potential faults that are not generally present. For instance, undetectable GNSS satellite faults can occur at rates greater than 10-7 per hour, and so must be accounted for in the PL.
In non-clear-sky environments, the GNSS position calculation is complicated by frequent loss of “sight” of the GNSS satellites. This is mitigated by having additional constellations and frequencies. However, for added availability of a precise position in challenging environments, it is necessary to incorporate sensor fusion into the position calculation, typically by using a six degree-of-freedom inertial measurement unit (IMU) as input, which includes three accelerometers and three gyroscopes to measure 3D translational and rotational motion. The IMU can maintain position accuracy for short periods when GNSS is unavailable, such as when driving under an overpass on a highway. The IMU provides a relative positioning output, so the absolute error growth is unconstrained in the absence of GNSS inputs. Therefore, it is important to have the GNSS receiver as the primary sensor in the positioning solution to constrain IMU drift and to reacquire GNSS signals rapidly after emerging from a GNSS outage.
Position error results for a typical highway environment are shown in FIGURE 8 after adding input from an automotive-quality IMU to the Hexagon PI software positioning engine. Small spikes in position error are due to short GNSS outages along the test route. However, the error growth due to loss of GNSS is minimal due to the coupling of the IMU data with the GNSS measurements.
FIGURE 8. Horizontal error time series and cumulative distribution function (CDF) of the TeseoAPP alone, and of the TeseoAPP with Hexagon PI software positioning engine (SWPE) in a highway environment. (Image: Authors)
FIGURE 9 shows the Hexagon PI highway data with accompanying PLs. Though the errors are well-constrained through GNSS outages, the PLs typically increase significantly. This is due to the higher noise of low-cost IMUs, and the uncertainty associated with reacquiring GNSS signals. PLs must account for worst-case IMU performance, which can have errors orders of magnitude greater than the nominal performance. During GNSS signal reacquisition, minimizing receiver noise is critical for fast position reconvergence, reinforcing the need for high-quality GNSS in autonomous applications.
FIGURE 9. Horizontal error and protection level (PL) including cumulative distribution functions (CDFs) of the Hexagon PI software positioning engine (SWPE) in a highway environment. (Image: Authors)
CONCLUSION
The TeseoAPP is the first generation of multi-band GNSS chipsets designed by STMicroelectronics to meet the two main requirements of autonomous driving: accuracy and safety-critical operation. The execution of the ISO 26262 standard for TeseoAPP is still a work in progress and encompasses two main aspects: 1) a safety plan implementation, code quality metrics and processes management and 2) the technical safety concept. Both of these aspects presented specific challenges, mainly due to the inherent complexity of the product and the large amount of firmware involved.
To exploit the maximum benefit of the TeseoAPP in safety-critical automotive applications, a high-accuracy ASIL-B-rated position engine is required. Hexagon PI’s software positioning engine is designed to use measurements from an ASIL-rated GNSS receiver, along with GNSS corrections and IMU data, to generate ASIL-rated position outputs, with accompanying integrity guarantees. The Hexagon PI software positioning engine computes protection levels. The calculation and determination of PLs is required to meet the safety and integrity guarantees necessary in autonomous driving for functionally safe operation.The software positioning engine also outputs ASIL-rated velocity, attitude and absolute time data, although we have not discussed these in this article.
The required high performance and safety expectations suggested, since the early stages of the project, a system composition in which the TeseoAPP was configured as an ASIL-B measurement-engine whereas the ASIL-B software positioning engine algorithms (by Hexagon PI) run on a separate ASIL host processor. We believe this synergy of competencies will represent the key for a successful solution to enable safe and reliable positioning in autonomous driving applications.
ACKNOWLEDGMENTS
The TeseoAPP chipset has been developed with the support and in the framework of the European Safety Critical Applications Positioning Engine project, which is funded by the European GNSS Agency under the European Union’s Fundamental Elements research and development program.
FABIO PISONI leads the GNSS System Architecture and Software Team (Automotive and Discrete Group) at STMicroelectonics Italy in Milan, where he has worked since 2009. He has a degree in electronics from Politecnico di Milano and has previous experience as a GNSS and digital signal processing (DSP) engineer.
DOMENICO DI GRAZIA is a GNSS signal senior staff engineer at STMicroelectronics Italy in Naples, where he has worked since 2003. He has a degree in telecommunication engineering from the University of Naples Federico II, holds patents in the GNSS area, and has previous experience in digital communications.
GIUSEPPE AVELLONE is in the GNSS System Architecture and Software Team (Automotive and Discrete Group) at STMicroelectonics Italy in Catania, where he has worked since 1998. He has a degree in electronics from Università di Palermo and previous experience as a GNSS and DSP engineer.
LUIS SERRANO is a GNSS technical marketing manager with STMicroelectronics based in Munich. He holds a Ph.D. in GNSS from the Department of Geodesy and Geomatics Engineering, University of New Brunswick, Canada. He has been active in the GNSS precise positioning field since 2007, and holds a patent on GNSS.
BRETT KRUGER is a software engineer specializing in GNSS/INS integration in the Safety Critical Systems Group at the Hexagon Positioning Intelligence (PI) NovAtel brandin Calgary, Canada. He holds an M.A.Sc. in electrical engineering from the University of Toronto, Canada.
LAURA NORMAN is a geomatics engineer specializing in GNSS integrity and protection levels in Hexagon PI’s Safety Critical Systems Group. She obtained her M.Sc. from the Department of Geomatics Engineering at the University of Calgary, Canada.
NATASHA WONG KEN is the Safety Critical Systems product manager at Hexagon PI. She has worked at Hexagon PI since 2012 after obtaining a B.Sc. in geomatics engineering from the University of Calgary.
The ZED-F9K turnkey solution minimizes the effort required to achieve decimeter-level positioning accuracy in automotive applications.
Siemens has integrated the u-blox ZED-F9K high-precision dead-reckoning module into its Toyota Prius V2X (vehicle-to-everything) test fleet. Siemens carried out live demonstrations of the technology at ITS European Congress 2019 in Eindhoven, the Netherlands.
As the only available source of absolute position, GNSS-based positioning plays a crucial role in advanced driver automation systems and driverless vehicles. The same is true in V2X communication, in which vehicles continuously share their location and other information with other traffic participants — cars and pedestrians — as well as surrounding infrastructure, improving road safety and reducing traffic congestion.
V2X test vehicles typically determine their position using high-end GNSS receivers. By opting to use the ZED-F9K, Siemens was able to align the performance of their test fleet with real-world conditions while also reducing the cost and the engineering effort required to develop their vehicles.
Siemens conducted V2X tests using the u-blox ZED-F9K during ITS European Congress 2019. (Photo: u-blox)
“We’ve had a very positive experience with u-blox’s ZED-F9K high precision dead reckoning solution. The product delivered strongly from the initial design-in to the data and performance in our first tests,” said Igor Passchier, engineering fellow, Connected and Automated Driving at Siemens PLM Software.
“Our collaboration with Siemens shows the extent to which the ZED-F9K turnkey solution saves OEMs time, cost, and engineering effort while providing decimeter-level positioning performance,” said Alex Ngi, Product Strategy for Dead Reckoning, Product Center Positioning, u-blox. “For us, it has also been a welcome opportunity to contribute to solving the challenges in the autonomous driving ecosystem.”
SoftBank plans to introduce a centimeter-accurate, real-time satnav-based positioning service, specifically using Japan’s Quasi-Zenith Satellite System (QZSS), to guide autonomous vehicles across a range of industries in Japan. The company said it will install more than 3,300 control points at base stations across Japan to deliver centimeter-level accuracy over its mobile network coverage area to provide real-time kinematic (RTK) positioning.
Testing begins in July with a scheduled launch of commercial service by the end of November. Test partners include Yanmar Agribusiness Co., Ltd., a provider of autonomous assisted driving for agricultural machinery, Kajima Corporation, which performs construction site management with automatically controlled drones for aerial photography and monitoring, and SB Drive Corp., a provider of autonomous and assisted driving technology for buses.
SoftBank is developing proprietary low-cost GNSS receivers so that “new services and market expansion can be realized.” A Positioning Core System provided by ALES Corp. will generate correctional data based on signals received and transmitted by SoftBank’s own control points over SoftBank’s mobile communications network to agricultural and construction machinery, self-driving cars, drones and other equipment carrying GNSS receivers. The company expects that centimeter-level positioning can thus be done in real time.
In addition to control points at its own base stations, SoftBank will use the Geospatial Information Authority of Japan’s approximately 1,300 GPS-based control stations.
SoftBank is also developing services to enablec loud-based RTK positioning for devices without GNSS receivers. Cloud-based RTK will provide centimeter-level, location-based services for equipment that needs to be miniature and energy-efficient, such as infrastructure surveillance sensors and wearable devices.
SoftBank Group Corp. is a Japanese multinational conglomerate holding company headquartered in Tokyo. It owns operations in broadband, fixed-line telecommunications, e-commerce, internet, technology services, finance, semiconductor design and more. It is the 36th largest public company in the world, and the 2nd largest in Japan.
ALES is a joint venture established by SoftBank and Enabler in July 2018. Enabler employs GNSS and related technologies to produce such products/services as a synchronization solution for mobile base stations for subway stations and a patented indoor positioning/time synchronization infrastructure platform in Japan.
A European Union project has designed and prototyped the ESCAPE GNSS Engine (EGE), a positioning module intended to enable autonomous or semi-autonomous driving functions.
Automated vehicles are on the way, and the European GNSS Agency (GSA) sees satellite navigation as a core technology that will help to ensure their safe operation. At the recent Mobile World Congress in Barcelona, the GSA shared its space with the ESCAPE project, an EU-funded initiative that has developed a unique positioning module for autonomous or semi-autonomous driving.
Autonomous vehicles will feature both sensor-based and connection-based solutions for a variety of vehicle services. Ultimately, the GSA sees a “converged solution” as the best alternative, combining the strengths of both approaches. By integrating sensor data and connectivity-based information, operators can reduce the need for the most expensive sensors and at the same time save money on infrastructure.
The Fundamental Elements-funded ESCAPE project has designed and prototyped the ESCAPE GNSS Engine. It is a unique positioning module that combines precision GNSS and 4G connectivity, for the highly accurate and reliable positioning capabilities required to make automated driving a reality.
The ESCAPE GNSS Engine. (Photo: GSA)
“This is an onboard unit for autonomous vehicles,” said Jessica Garcia Soriano, R&D engineer of the Advanced Communications Business Unit at Ficosa. “It is equipped with a very good GNSS receiver made by STMicroelectronics. This was actually the first dual-frequency GNSS receiver made for the automotive market.”
Dual-frequency is of course a real differentiator for Galileo, as the world’s leading provider of dual-frequency GNSS signals. This means added precision and robustness and it helps enormously with multi-phase errors and other urban canyon issues in city-driving scenarios.
“We also have a very good positioning solution provided by GMV, another Spanish company. They are experts in these kinds of solutions. The outputs from this solution are very accurate. So we have GNSS of course, including Galileo, and apart from this you have a modem inside, a 4G modem that gets GNSS corrections from the internet, so this helps to provide better positioning. And apart from this you have inside the same module an inertial measurement unit [IMU]. This is a sensor, a device that senses acceleration and has a gyroscope, so this information also helps in providing good positioning.”
The ESCAPE unit also provides for the integration of other data from the vehicle. “That means vehicle odometry, for instance, you can have camera information, or information from maps that are stored in the vehicle, among others” Garcia said.
The market is ready
“One of our important goals is to provide a low-cost system,” Garcia said. “There are other very good positioning systems that are being developed that can be based on some very advanced technologies, such as LiDAR for instance, but this is very expensive. So our target is to develop and build a prototype of a system that could be installed in all vehicles, for the whole market. And so we are combining GNSS, 4G, IMU and all of these other data sources from the vehicle in an intelligent way, in an affordable way.”
Indeed, one of the things that make ESCAPE unique is the way it brings together high-end GNSS processing capabilities with an industrialisation process that targets high volumes and comparatively limited cost and size. It also encompasses hardware and software safety procedures required for certification for the automotive market.
Garcia explained, “At Ficosa, we are a top-tier global provider devoted to the research, development, manufacturing of vision, safety, connectivity and efficiency systems for the automotive sector. We provide solutions directly to vehicle manufacturers. Based on our expertise and thanks to the work we have done on this project, we understand very well that GNSS is a central focus for a lot of applications. From the moment we started working on this project, at Ficosa we realised that this is a new and very important market. Right now we are working on a positioning system for autonomous driving based on this unit. This is part of our roadmap at the moment. This is a positioning system that we are ready to offer to the customer.”
The unit is ready now, but we have yet to see autonomous cars in large numbers on the road. Is this a problem for the ESCAPE system? Garcia answered, “From the very first moment that you have an autonomous car in the street, you will need high-accuracy positioning, because these vehicles will need this positioning to maintain themselves safely on the road. But we don’t have to wait for autonomous cars. The vehicles on the road today can already benefit from this technology.”
Garcia pointed to Europe’s eCall system, where a call centre automatically receives location information from vehicles in distress, thanks to on-board GNSS. “You already have this emergency call technology in the vehicles,” Garcia said, “and it provides a location, so the better the location is, the easier it is to locate the people in an emergency situation. No, we don’t have to wait.”
Location and more
One thing everyone seems to agree on is that autonomous vehicles will soon be appearing on European road networks, and most driving-related decisions will be based, one way or another, on the location of the vehicle and of other vehicles and objects in its vicinity. So vehicle location and positioning will be a critical component for the effective transportation of people and goods by self-driving road vehicles. That positioning will be enabled mainly by GNSS technologies, including Europe’s Galileo, which is expected to offer significant benefits in terms of accuracy and authentication compared to the other satellite-based navigation systems.
GNSS-based location will have to be complemented by other technologies in order to get to the integrity level needed in all driving situations, but the GSA also believes the combination of dual-frequency GNSS and 4G/5G connectivity can do more than just navigation, enabling as well a diverse range of in-vehicle location-based services (LBS), much like what we see emerging in smartphones. The EU-funded ESCAPE project, with its innovative GNSS engine, represents an important step forward in the pursuit of accurate, reliable and affordable positioning and connectivity for the emerging autonomous and connected cars markets.
By Daniela E. Sánchez, Harvey C. Gómez and Thomas Pany, Institute of Space Technology and Space Applications (ISTA)
This paper presents how our system, consisting of a GNSS receiver antenna, an inertial measurement unit (IMU) and a lidar, is used to obtain high-precision maps through the geo-referencing of lidar point clouds. An accuracy assessment of the system is conducted, which also gives us insights on the quality of lidar range measurements for autonomous driving applications.
The assessment is done by geo-referencing the obtained point clouds of extracted buildings and comparing them against a supporting measuring system like a total station. The building extraction is done by performing an approximation of the mathematical model of a plane to the facades that composes the building in both, the lidar and the supporting measurement system data.
The paper also indicates the proposed pose determination method of a mobile agent using lidar data. Thanks to the advantages of active, 3D sensors, diverse objects in the environment can be detected as individual point sets, or clusters. Each of the segmented objects can be used as a landmark to figure how the agent is located with respect to those structural elements. The algorithm is capable of detecting the clusters in one point cloud, and finding the most alike point set on a subsequent scan. This is achieved by comparing global descriptors for point cloud data.
The Ensemble of Shape Functions (ESF) is selected as the cluster descriptor. The cluster matching is performed by comparing the clusters one-to-one, calculating the minimum Chi-squared distance among their descriptors. The smaller this distance, the greater the probability of being the same cluster in distinct epochs.
Figure 2. Direct geo-referencing of lidar data at different times. (Image: Authors)
The resultant cluster correspondences for the whole point cloud allow finding the rigid transformation between the point clouds. An initial coarse alignment among the clouds based on the centroids of each matched cluster was performed, followed by a fine alignment in order to reduce errors by the use of the Iterative Closest Point (ICP) algorithm. This approach is valid for urban environments, or for those where many objects can be segmented as clusters.
Finally, a practical case is described in order to show how we plan to use the outcome of the highly precise geo-referenced point clouds and the pose estimation method using lidar.
“As of today [Nov. 28] Tesla owners have driven 1 billion miles with Autopilot engaged,” the company announced via tweet.
The Autopilot feature became available in 2015 and now comes on all new Tesla models with a $5,000 activation fee at the time of purchase or $7,000 if selected later.
The company is training its “neural networks” to improve its self-driving system.
Photo: Tesla
Tesla’s global fleet totals more than half a million vehicles, and recently marked a 20-billion mile step of total electric miles driven, the company said.
The Autopilot system can also function in the background of the vehicle, without being activated and with no input on control. Thus it gathers data from many more billions of “drivered” miles about its environment and potential Autopilot behavior.
The company previously mentioned the 1 billion-mile autonomous mark as the minimum it would need to move Autosteer from beta to a regular feature.
Updates to Autopilot are planned for 2019, including new hardware that will aid in the rollout of the company’s Full Self-Driving system, possibly by the end of that year.
The U.S. Department of Transportation’s most recent document preparing for the future of self-driving cars almost entirely ignores positioning, navigation and timing (PNT) needs, according to the Resilient Navigation and Timing (RNT) Foundation. And when it does address GPS, it gets things wrong. A Dec. 3 deadline looms for interested parties to file their comments with DOT.
The comment period is open until December 3. Interested parties can make their own comments and read those already submitted at the website for Docket DOT-OST-2018-0149.
The cited comment from the RNT Foundation states that, while most self-driving cars are being designed to navigate without external inputs, GPS/GNSS will still be required to initialize location information for vehicle cold startups. Also, most vehicles will reference GPS/GNSS when communicating their positions to other vehicles and traffic control systems.
Much of the benefit of automated vehicles will come from their participation in Intelligent Transportations Systems. This means wireless networks. The RNT Foundation also urges the department to consider these networks’ critical dependence on GPS timing synchronization in their plans going forward.
(Image: Pavel Vinnik/Shutterstock.com)
The Secretary of Transportation has had a mandate to provide a backup capability for GPS since 2004 that has not been acted upon. The RNT Foundation comments observe that doing so could greatly mitigate all of the concerns mentioned.
Dana Goward is president of the Resilient Navigation and Timing Foundation, based in Washington D.C.
Quectel Wireless Solutions has launched the automotive-grade C-V2X module AG15, which features the Qualcomm 9150 C-V2X chipset solution from Qualcomm Technologies, a subsidiary of Qualcomm Incorporated.
Image: Quectel
The module is now sampling to the automotive industry for the development of commercial C-V2X products.
The AG15 C-V2X module is manufactured in accordance to IATF 16949:2016 quality management system standard for the automotive sector, and it fully complies with the automotive product quality processes, including APQP and PPAP.
Pairing with the Quectel automotive-grade LTE module AG35, Quectel’s AG15 is designed to meet the telematics and vehicle-to-everything (V2X) connectivity requirements of the next-generation automotive applications, such as autonomous driving and road safety.
Also known as LTE-V2X, C-V2X is the V2X communication technology based on the globally recognized Third Generation Partnership Project (3GPP) Release 14 specifications. The PC5-based direct communication mode of C-V2X supports vehicle-to-vehicle (V2V), vehicle-to-infrastructure (V2I) and vehicle-to-pedestrian (V2P) communications on the 5.9-GHz intelligent transport system (ITS) spectrum.
In addition, C-V2X paves a strong evolution path toward 5G new radio (5G NR) and plays an essential part of the future of safe autonomous driving with its capabilities including non-line-of-sight (NLOS) sensing to support high-speed mobility and high vehicular density deployments, the company said.
For positioning function, the AG15 features a built-in multi-constellation high-precision GNSS (GPS/GLONASS/BeiDou/Galileo/QZSS) receiver, with additional support from satellite-based augmentation systems (SBAS) and Qualcomm 3D dead-reckoning technology, which greatly improves the positioning accuracy and speed while simplifying customer designs.
Based on C-V2X technology, the Quectel AG15 module adopts the 3GPP Release 14 C-V2X PC5 protocol. It is designed to allow low-latency, highly reliable and highly dense data exchange between vehicles and their surroundings, enabling effective information sharing among road users in avoidance of collisions, thus improving automotive safety, automated driving and traffic efficiency.
Without the need for a subscriber identity module (SIM), cellular subscription or network assistance, the C-V2X direct communication mode helps reduce complexity and cost for customers.
Additionally, the Qualcomm 9150 C-V2X chipset solution has a built in A7 application processor (1.5 GHz), which could be potentially utilized to run ITS stack and associated C-V2X applications.
“We are very pleased to introduce our first batch of automotive grade C-V2X modules based on the Qualcomm 9150 chipset solution. Automated driving has unique requirements for V2X connectivity, such as lower latency, higher reliability and wider bandwidth, all of which could be addressed by C-V2X technology,” said Patrick Qian, CEO of Quectel. “Built upon Quectel’s expertise in connected vehicles and Qualcomm Technologies’ high-performance C-V2X chipset solution, the AG15 module is expected to help automakers and Tier 1 suppliers to effectively accelerate their efforts towards automated driving.”
“Quectel is a global leader in cellular modules with rich experience in commercial automotive products,” said Nakul Duggal, vice president of product management, Qualcomm Technologies. “We are pleased to work closely with Quectel again to support its modules with our 9150 C-V2X chipset solution to help create safer and more efficient V2X solutions and to help empower road safety and autonomous driving. We look forward to working with Quectel in delivering the solutions required to address the needs of the next generation automobiles.”
Israel-based Autotalks has launched what it calls a global V2X (vehicle-to-everything) chipset.
The chipset supports both dedicated short-range communications (DSRC) and cellular vehicle-to-everything (C-V2X) technology — both allow vehicles to share their location and speed to help prevent accidents and improve the safety of autonomous driving systems, the company said.
The chipset’s processor also could allow customers to switch between the two standards. It minimizes development, testing and certification efforts for a V2X system to be deployed anywhere via a software-defined toggle between the two V2X technologies.
Two competing standards
Automakers have announced intentions to equip their new car models with V2X technology. In recent years, V2X has diverged into two different solutions, DSRC and C-V2X.
While DSRC-based V2X is deployed in the U.S., Europe and Japan, C-V2X is gaining momentum in other regions. Its fundamentally different architectures have made it difficult to harmonize a single global solution.
Autotalks’ response is to equip its second-generation chipsets with C-V2X in addition to native support of DSRC.
Autotalks’ deployment-ready, second-generation V2X chipset supports both DSRC and C-V2X direct communications (PC5 protocol) at the highest security level. According to the company, the chipset supports DSRC based on 802.11p/ITS-G5 standards and C-V2X based on 3GPP specifications.
Autotalks said its chipsets were designed to meet V2X market requirements and standards, including security, environmental, quality, thermal and other requirements.
Hexagon’s Positioning Intelligence division has achieved a milestone toward its goal of safe autonomy on the road. The division is developing functionally safe positioning technologies for fully autonomous vehicles and other applications.
A third-party audit has been completed that confirms process compliance with key automotive specifications ISO/TS 16949 and ISO 26262 Functional Safety Design Assurance. This is an important step toward the development of functionally safe new technology that meets the exceptional safety standards set by the automotive industry, Hexagon said.
“We’re thrilled to have our core engineering processes updated to meet the requirements of automotive applications,” said Jonathan Auld, vice president of Safety Critical Systems, Hexagon’s Positioning Intelligence division. “We are building on a 25+ year history in safety of life solutions for the marine and aviation industries, and we expect this leadership to serve us well in automotive.”
NovAtel has demonstrated high-accuracy positioning performance using automotive-grade GNSS chipsets Teseo APP and Teseo V from STMicroelectronics. Combining automotive-grade multi-frequency GNSS chipsets with positioning algorithms and correction services from NovAtel improves the achievable positioning accuracy available to automotive users and provides a solution suitable for autonomous operation.
According to the company, these chipsets provide multi-frequency GNSS data for precise point positioning (PPP) and real-time kinematic (RTK) to enable accurate positioning capabilities. Teseo APP features built-in integrity checking for use in safety-critical systems, whereas Teseo V is used for non-safety-critical precise positioning applications.
The collaboration between the two companies is designed to reach car manufacturers and Tier 1 suppliers for future production models.
Test results: Horizontal position errors. Teseo V alone is shown in red, Teseo V + NovAtel in green. Test results: Horizontal cumulative error distribution. Teseo V alone is shown in red, Teseo V + NovAtel in green. (Chart: NovAtel)
Test results: Horizontal cumulative error distribution. Teseo V alone is shown in red, Teseo V + NovAtel in green. (Chart: NovAtel)
Driven Today. “STMicro is one of many chipset manufacturers coming to market with dual-frequency chipsets targeting the automotive sector,” said Jonathan Auld, VP Engineering and Safety Critical Systems for NovAtel. “We are taking advantage of their expertise in automotive measurement engines for high-volume, cost-effective reliable positioning. NovAtel brings high-precision algorithm expertise and integration with global corrections supplied by Hexagon Correction Services to this initiative.”
NovAtel’s positioning engine combines the GNSS measurements from these chipsets with inertial measurement unit (IMU) data and Hexagon Correction Services to deliver centimeter-level PPP positioning solutions in real time.
“Working closely with STMicroelectronics allowed us to innovate and drastically reduce time to market of our assured positioning solution tailored specifically for safe positioning of autonomous vehicles,” added Auld.
Comparison of GNSS Performance possible in automotive today (red), L1 automotive with corrections (green) and L1/L2 automotive with corrections (blue).
Driverless Tomorrow. “Precise absolute positioning is just one piece of the overall autonomous vehicle puzzle and must be done with safety and integrity concepts in mind.” Auld pointed to the partnership announced in 2016 between NovAtel, the Illinois Institute of Technology, and Stanford University to conduct leading-edge research to determine how GNSS technology can deliver a positioning solution that meets both the safety and accuracy requirements of autonomous automotive vehicles.
Previous research by academia and industry into GNSS integrity produced the successful WAAS program for aviation. The new work underway will extend the scope to include the autonomous ground vehicle use case. The research includes updated and expanded concepts for high-integrity carrier-phase algorithms as well as expanded threat models and safety monitors.
At the Automotive Tech.AD in Berlin, Auld added: “Today the primary use case for positioning in navigation is single-frequency GNSS, with up to 2 constellations, using narrowband RF and antennas, obtaining accuracy at the 1–2 meter level. This is primarily done with pseudorange-based positioning techniques, with some carrier-phase assistance. There are no functional safety standards, and so safety data is provided on the output solution.”
Autonomous Requirements. By contrast, he continued, autonomous operation will require lane-level and better accuracy: 3D centimeter to decimeter absolute positioning. This means multi-frequency, multi-constellation receivers and antennas to improve overall accuracy and increase available measurements. It will also require increased availability through sensor fusion with IMUs and other sensors. All of this must be brought together through a functionally safe development process targeted at ISO26262 Automotive Safety Integrity Level (ASIL) B.
Moving from meter to centimeter level position requires additional processing to handle all the added signals coming in; residual monitoring and observation exclusion, and carrier phase, “the key to centimeter-level positioning,” as opposed to code phase. The vehicle’s localization system must include enhanced positioning algorithms for multipath mitigation, a fast converging corrections network, enhanced Kalman Filters, and sophisticated sensor fusion.
Flexible Integration. NovAtel’s positioning engine architecture enables a flexible integration with different GNSS receiver chipsets, augmentation sensors and processor environments, providing automotive manufacturers with additional flexibility when it comes to sourcing of components and subsystems of advanced driver assistance systems (ADAS) and autonomous driving solutions.
The positioning engine is being developed to ASIL-B standards and will include a proprietary GNSS integrity solution to ensure safe positioning within defined protection limits tailored to the customer’s application requirements.
