Publicerad den Lämna en kommentar

Estimating heights with subsidence changes using NGS data and tools

This column details the potential effects of crustal movement on published heights in various regions of the United States.

In my last column (in the April 2021 Survey Scene), I mentioned that the National Geodetic Survey (NGS) announced that it is suppressing height information in Southeast Texas.

The April column also highlighted one of NGS’ four use cases – “Use Case 1: Flood Mapping.” The case study discusses the Elevation Certificate (CE) Example, Flood Insurance Rate Map (FIRM) and Flood Insurance Study (FIS).

The column highlighted the potential effects of subsidence on published heights in the Houston region, which implied that most of the published heights that are based on older surveys in the region are not current or accurate.

This column will provide more details of the suppression of heights in the Southeast Texas region, and potential effects of crustal movement on published heights in other regions of the United States.

NGS announcement to suppress height information for Southeast Texas. (Image: NGS)

NGS announcement that it suppressed height information for Southeast Texas. (Image: NGS)

According to NGS’ announcement, only 28 marks will have publicly available orthometric heights on NGS datasheets in Southeast Texas.

The “Link to Map: SE TX Valid Ortho. Heights” button provides the benchmarks available to users (see the box titled “Link to Map SE TX Valid Ortho Heights”). The website provides links to the published stations.

Clicking on an icon provides the PID and name of the station with a link to a datasheet. Click  “Get Datasheet” for a datasheet of the station. Below is an excerpt from the datasheet of Station P 1200.

Let’s address why NGS is suppressing the stations in Southeast Texas. My last column provided plots depicting the amount of movement in the Harris-Galveston, Texas, region. See the box titled “Estimate of Amount of Subsidence in 5 Years in Harris-Galveston, Texas, Region – Units Feet.”

As indicated in the plot, some of the marks are estimated to have moved almost ½ foot (approximately 0.15 meters) in 5 years. In addition, some of the relative height differences approach 1/3 of a foot (approximately 0.1 meter) between neighboring stations. See the highlighted stations in the box titled “Estimate of Amount of Subsidence in 5 Years in Harris-Galveston, Texas, Region – Units Feet.”

The last major releveling incorporated into NGS’ Database in the Harris-Galveston, Texas, region was performed more than 30 years ago in the 1986/1987 timeframe. Therefore, some of the published stations in the region could have subsided more than three feet (or about a meter).

As stated in NGS’ Blueprint 3, “Most leveling data in NGS archives comes from the mid-20th century, in support of the NAVD 88 project.” Of course, most regions of the United States are not subsiding at the same rates as in the Houston-Galveston, Texas, region.

In a previous newsletter, I discussed NGS’ second Multi-Year CORS Solution of the National CORS (MYCS2). I downloaded the coordinates and velocities from NGS’ website and created a plot of the vertical velocities. For those who prefer to use feet as opposed to meters, I provided velocities with units in feet/year and mm/year.

See the boxes titled “Estimate of Velocity Rates Based on MYCS2 – CONUS (feet/year),” “Estimate of Velocity Rates Based on MYCS2 – Alaska (feet/year),” “Estimate of Velocity Rates Based on MYCS2 – CONUS (mm/year)” and “Estimate of Velocity Rates Based on MYCS2 – Alaska (mm/year).”

It should be noted that the intent of these four plots is to provide a wide-ranging view of the values and some of the variation in rates across the United States.

Estimate of Velocity Rates Based on MYCS2 – CONUS (feet/year). (Image: David Zilkoski)

Estimate of Velocity Rates Based on MYCS2 – CONUS (feet/year). (Image: David Zilkoski)

Estimate of Velocity Rates Based on MYCS2 – CONUS (feet/year). (Image: David Zilkoski)

Estimate of Velocity Rates Based on MYCS2 – CONUS (feet/year). (Image: David Zilkoski)

Estimate of Velocity Rates Based on MYCS2 – CONUS (mm/year). (Image: David Zilkoski)

Estimate of Velocity Rates Based on MYCS2 – CONUS (mm/year). (Image: David Zilkoski)

Estimate of Velocity Rates Based on MYCS2 – Alaska (mm/year). (Image: David Zilkoski)

Estimate of Velocity Rates Based on MYCS2 – Alaska (mm/year). (Image: David Zilkoski)

The rates appear to be small in most regions of the United States. As an example, the rates are all less than -0.0062 feet/year (-0.0019 meters/year) in the Lake Norman region in North Carolina (see the box titled “Potential Subsidence Rates in the Lake Norman Region in North Carolina). It would take many years for the crustal movement to make a difference to some projects in this region.

Potential Subsidence Rates in the Lake Norman Region in North Carolina. (Image: David Zilkoski)

Potential Subsidence Rates in the Lake Norman Region in North Carolina. (Image: David Zilkoski)

That said, let’s look at another region of the country. For example, in the vicinity of Maryville, Missouri, the rate of subsidence is around -0.0187 feet/year (-0.0057 meters/year). See the box titled “Potential Subsidence Rates in the Maryville, Missouri, Region.” These subsidence rates don’t appear to be large values but if you take into account the last time the height of a mark was established by leveling data it could result in a large difference from the true orthometric height.

Potential Subsidence Rates in the Maryville, Missouri, Region. (Image: David Zilkoski)

Potential Subsidence Rates in the Maryville, Missouri, Region. (Image: David Zilkoski)

According to NGS’ database, it appears that many of the marks in the Maryville, Missouri, region were last leveled in 1935. I used NGS’ Passive Mark Lookup tool and Leveling Project Page tool to identify the marks and associated leveling lines in the area of the CORS stations in the Maryville, Missouri, region.

I described the Passive Mark Lookup webtool in a previous column. As previously mentioned, these subsidence rates all seem very small, but if you take into account the last time the height of mark was established by leveling data, the subsidence value can be very large.

See the box titled “Potential Subsidence in 86 Years in the Maryville, Missouri, Region.” The box indicates that, if you account for the last 86 years (2021 – 1935), the potential subsidence exceeds 1½ feet (-1.6082 feet, -0.4902 meters).

Potential Subsidence in 86 Years in the Maryville, Missouri, Region. (Image: David Zilkoski)

Potential Subsidence in 86 Years in the Maryville, Missouri, Region. (Image: David Zilkoski)

Continuing across the country to Colorado, the box titled “Potential Subsidence Rates in the Grand Junction Region, Colorado,” provides the estimate of subsidence rates in Mesa County, Colorado. As the plot indicates, the rates vary between -0.0046 feet/year (-1.4 mm/year) and -0.0128 feet/year (-3.9 mm/year). Once again, these rates all seem relatively small but many of the marks near CORS MC06 were last leveled in 1985. This means the potential change in height could be as large as 0.2592 feet (0.0792 meters).

Potential Subsidence Rates in the Grand Junction. Colorado, Region. (Image: David Zilkoski)

Potential Subsidence Rates in the Grand Junction Region, Colorado. (Image: David Zilkoski)

Obviously, this is only an estimate of the subsidence in the region and the actual amount of subsidence is unknown since the last time the mark was leveled. These estimates are based on the MYCS2, which uses current data to estimate the velocity. The processing included data spanning 1996 to 2016 (week 0834 to 1933), 1099 weeks or about 21 years in total.

The point of this column is not to provide the exact change in height of a mark, but to highlight that the publicly available orthometric height on a NGS datasheet may not be up to date based on crustal movement. The new modernized National Spatial Reference System will enable users to determine an accurate, current height on a mark and be able to efficiently and effectively monitor changes in a mark’s height.

As stated in NGS’ announcement to suppress the heights in Southeast Texas, the agency has developed tools to assist users in submitted data to NGS. See the box titled “Excerpt from NGS Announcement to Suppresses Height Information for Southeast Texas.”

Excerpt from NGS Announcement to Suppresses Height Information for Southeast Texas. (Image: NGS website)

This assistance is for every user, not just for individuals performing surveys in Southeast Texas. NGS has Regional Geodetic Advisors throughout the United States.

NGS Regional Geodetic Advisors. (Image: NGS Website)

NGS Regional Geodetic Advisors. (Image: NGS Website)

The Regional Geodetic Advisors provide guidance and assistance to constituents within their region. They are subject-matter experts in geodesy and regional geodetic issues. These individuals can assist users that are planning GNSS campaigns to re-densify the network.

NGS also provides a website detailing how users can help densify the network to prepare for the new, modernized North American-Pacific Geopotential Datum of 2022 (NAPGD2022). See the box titled “NGS GPS on Bench Marks Webpage.”

As mentioned in previous newsletters, a benefit of the new modernized National Spatial Reference System (NSRS) will facilitate the establishment of consistent, accurate NAPGD2022 GNSS-derived orthometric heights.

NGS GPS on Bench Marks webpage. (Image: NGS Website)

NGS GPS on Bench Marks webpage. (Image: NGS Website)

This column provided details on the suppression of heights in the Southeast Texas region, and potential effects of crustal movement on published heights in other regions of the United States. NGS suppressed the heights in the Southeast Texas region because of the large amount of crustal movement since the last time the heights of the marks were established.

As indicated by NGS’ MYCS2 velocities, every mark could be affected by crustal movement. In my opinion, the question a user should be asking is “How much has the height of the mark changed since it was last determined? Not, “Has the height of the mark changed?”

Publicerad den Lämna en kommentar

Editorial Advisory Board PNT Q&A: GNSS diminishing returns?

As the number of GNSS constellations and satellites in orbit continues to grow,
will we reach the point of diminishing returns?

Ellen Hall

Ellen Hall

“More satellites equal more data, and redundant constellation systems — through GNSS interoperability — can give us more robust PNT, as restated in the January Memorandum on Space Policy Directive 7. That said, there are always diminishing returns. Treaties place liability on the launching country if something goes wrong, but with tens of thousands of small satellites expected to be launched over the next decade, it will be getting increasingly crowded. Concerns are growing about the necessity of increased maneuvers to keep these satellites from a chain reaction of collisions, which ultimately could cause debris to fall to inhabited areas of Earth.”
— Ellen Hall / Spirent Federal Systems

Jean-Marie Sleewaegen

Jean-Marie Sleewaegen

“With already more than 130 GNSS satellites in orbit, the benefit of new satellites decreases while the risk of satellites interfering with each other increases. However, this is only considering GNSS as we know it, in the MEO orbit (altitude about 22,000 km). The future of GNSS may well be closer to Earth, in the LEO orbit (<1,000 km), with well-known benefits in terms of convergence time and resilience to jamming. Sooner than later, we can expect constellations of hundreds or thousands of LEO satellites carrying a GNSS-like payload supporting PNT services. No worries, there is still growth potential!”
— Jean-Marie Sleewaegen / Septentrio

Headshot: Stuart Riley

Stuart Riley

“With the current four GNSS constellations and a typical survey elevation mask of 10˚in North America, we average around 30 visible satellites. Far more are visible in Asia with the addition of the regional systems. In an area with a clear view of the sky, this provides more than enough satellites for precision centimeter positioning. However, most professional GNSS users do not have the luxury of operating exclusively in open areas with ideal conditions. Accessing many satellites across multiple constellations increases the probability of receiving sufficient satellites that produce high-quality measurements in obstructed areas. As the constellations expand, we observe improvement in precision position availability in these locations. The large number of satellites, coupled with independence across the four systems, improves system integrity and continuity while also helping to reduce the converge time in PPP solutions.”
— Stuart Riley / Trimble

Bernard Gruber

Bernard Gruber

“In a utopian vision of navigation, data gluttons and like-users of GNSS would say that there will never be enough! If capabilities remained static, then yes, I believe we would reach the point of diminishing returns. I would offer that innovation and competition will continue to drive capability improvements via power, signal quality, coverage, integrity and clock/timing accuracy. These innovations, coupled with user equipment flexibility utilizing signals from space, will drive an ever-maturing market balance and increasing return.”
— Bernard Gruber / Northrop Grumman

Publicerad den Lämna en kommentar

First Fix: Let’s start a conversation

Matteo Luccio

Matteo Luccio

Last year, GPS World marked its 30th anniversary. That is a testimony to this magazine’s continued relevance, its commitment to its marketing partners, and its unmatched audited audience of 54,000 GNSS/PNT buyers, integrators and specifiers.

Taking on my new role on GPS World’s edit team was a homecoming of sorts because I began my current career a little more than 20 years ago as this magazine’s managing editor. I look forward to an ongoing conversation with many of you in the GNSS/PNT community — scientists, engineers, civil servants, uniformed service members, company executives and product managers. You may get an email message from me, and I will always welcome yours, at the e-mail address below.


“I look forward to an ongoing conversation with many of you in the GNSS/PNT community.”


Let me tell you a little about three passions that led me to this job.

Navigation has been one of my passions since I was a kid. When I was five years old, I lost track of my mother as she entered a store in Berkeley, California, and I kept walking down the street. It happened again when I was seven and had insisted on walking home alone from school in Milan, Italy. I was determined never to get lost again. So, when I was 11 and my family moved to Pisa, I was the only kid I knew who walked around — from school to sabre-fencing practice, to piano lessons, to my bus stop — studying a map and a compass. When I was 13, in shop class, I built a crude optical-range finder, based on trigonometry. Next, came the topo maps I used for hiking the hills and mountains of Tuscany. A few years later, as a graduate student at MIT, I began to sail around the Boston Harbor islands and off the coast of Maine. I learned to navigate using nautical charts, sextants, radio direction-finders, sonar, radar, Loran-C and, finally, GPS receivers.

Magazine journalism has been another one of my passions, since I co-founded a public policy magazine, Oregon’s Future, 25 years ago and became its editor. That was the first of seven editorial positions with magazines I have had over the past quarter century. Finally, my passion for public policy led me to degrees in political science and to my previous career as a research analyst — first for an independent research institute, then for state and local government. It gave me a solid grounding in public policy, statistical analysis and querying large databases.
So, this navigation enthusiast, policy wonk and experienced writer and editor is now in position to report on and advocate for the continuing growth and development of GNSS. I will also showcase new products and projects, and present facts and opinions from across the GNSS/PNT community.

This month’s cover story on autonomous vehicles is a perfect example. It confirms the central role of GNSS in one of the most significant technological advances we can expect to see in the coming decade — having vast implications for our society and environment — with facts and opinions from four industry leaders.

Matteo Luccio | Editor-in-Chief
mluccio@northcoastmedia.net

Publicerad den Lämna en kommentar

GNSS in the fast lane: Meeting autonomous vehicle navigation challenges

Autonomous vehicles are being tested both on open roads and in controlled environments. (Photo: Trimble)

Autonomous vehicles are being tested both on open roads and in controlled environments. (Photo: Trimble)

The advent of autonomous vehicles (AVs) is one of three revolutions in the automotive industry that will likely change this country as much as cars did over the last century. The other two are the conversion from internal combustion engines to electric ones and the integration of cars into digital traffic networks.

Once mass deployed, AVs promise to dramatically reduce the number of traffic fatalities (42,000 in the United States in 2020, a National Safety Council report shows). They will never be sleepy, distracted, aggressive or drunk — nor will they engage in such inane human driving behaviors as texting while driving, playing chicken with bicyclists, or running red lights. They also promise to reduce fuel consumption, harmful emissions and traffic congestion by optimizing routes and increasing the number of people using car services instead of owning their own car.

To realize this vision, however, cars will have to do a lot more than just find their way on their own. They will have to perform flawlessly in an unpredictable world that includes toddlers, reckless drivers, fallen trees, sinkholes, construction and accidents.


See also Testing autonomous vehicles inside and out


GNSS Plus Corrections

Among the many sensors aboard an AV, the GNSS receiver has a unique role. It is the only one that can provide absolute positioning, in the form of latitude and longitude coordinates, to within a couple of decimeters anywhere on Earth. As such, it is “a key enabler to a lot of the vehicles to know precisely where they are and whether it is safe to activate autonomous systems,” says Gordon Heidinger, automotive segment manager, Autonomy and Positioning division at Hexagon.

A GNSS receiver cannot achieve the level of accuracy required for autonomous driving without robust corrections. Fifteen years ago, the state of the art was real-time kinematic (RTK) corrections. However, “the cost of that equipment exceeded the cost of a small car at that time,” recalled Steve Ruff, general manager, On-Road Autonomy Division at Trimble. “They were targeting a system cost of about $200. Today, that number is below $50, including the antenna, the GNSS positioning engine, and the software that runs on it.”

Today, all automotive manufacturers are using a form of precise point positioning (PPP) corrections, which is a one-way broadcast, as opposed to the two-way communication between a base station and a rover required for RTK. This means that a single correction stream can serve an entire continent, Ruff pointed out. “Once a vehicle is manufactured, we will support it with our PPP corrections stream for at least 10 years, which is the typical service life of a vehicle.”


Obstacles to Adoption

To achieve mass-market adoption, AVs will have to overcome numerous and complex obstacles:

  • The technical difficulty of dealing with a limitless number of unanticipated challenges, such as poor visibility because of weather conditions, unpredictable human behaviors, complicated obstructions, detours and potholes
  • The need to map millions of miles of roads, develop vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communications, and protect vehicle software from hackers
  • The difficulty, if not the impossibility, of handing off control to a human quickly enough to be safe when the system is unable to deal with a complex situation
  • Questions about legal responsibility and insurance liability
  • Ethical dilemmas about how to program the system to respond in emergencies
  • The development of appropriate federal and state regulations
  • Resistance from paid drivers who fear losing their jobs, including 3 million U.S. truckers, and from many other drivers, who fear losing control over their safety.

Trimble has approached all the major car manufacturers, has several programs in development, and has received multiple positioning requests for information (RFIs), Ruff said. “In 2018, Trimble’s RTX corrections service was the first solution adopted for production use in passenger vehicles, providing absolute precise positioning for General Motors’ Super Cruise system.”

Additionally, Trimble is working with Qualcomm and with SiriusXM, which will deliver Trimble’s RTX corrections over its satellite network, just like it does with music. “It is a good partnership because about 80% of the vehicles in North America are coming equipped with SiriusXM radio technology,” Ruff said. “The OEMs do not have to buy any additional hardware.” RTX corrections can also enter a vehicle via cellular IP, L-band satellite broadcasts and, potentially, over a V2I link.

Hexagon has proposed a PPP solution for automotive, “mainly because we essentially have the world covered with base stations, and that is a hard thing to do,” Heidinger said. “We have been running a corrections network for a very long time.” PPP’s one-way broadcast offers better cybersecurity because the GNSS receiver does not have to disclose its position, he added.

Swift Navigation is building a global corrections network. To make it suitable for the automotive market, the company is aiming to make its corrections service affordable and scalable. “We realized quickly that neither of the traditional RTK and PPP approaches were going to meet those requirements,” said Fergus Noble, company co-founder and CTO, “so we invested in developing a corrections service pretty much from the ground up.”

RTK provides high accuracy and short convergence times but is typically costly to deploy because it requires a very high density of stations, Fergus explained. As a consequence, most providers do not have continuous coverage over a wide area. Conversely, while PPP is a true global solution, it is less accurate and takes a long time to converge. “That may be fine in a marine or land-surveying application, but not if you are driving through city tunnels and bridges and need it to be able to reacquire a high-accuracy position within a matter of seconds. Therefore, we took a hybrid approach, together with a lot of new IP that we developed.” The service provides coverage in all the United States and most of Europe, and is being tested in Japan, South Korea and Australia.

Accuracy and Integrity

A common target accuracy for lane-level positioning is 20 cm 95% of the time. That means that AVs need to know when their positioning accuracy falls beneath that threshold. “We are building into our positioning solutions an accuracy metric that is output along with the position information we are providing,” Ruff said. “[The metric] can be used by the intelligence in the system to decide whether it can rely on the GNSS solution or needs to switch to one of the other complementary technologies because GNSS accuracy is not fulfilling its lane discipline.”

Heidinger noted the importance of economies of scale when mass-producing vehicles, where cost and ease of manufacturing become factors. “We can take some of our high-end equipment and get you 2 cm of accuracy with this technology, but the price point and the feasibility of this going into mass production for automotive is not favorable,” he said. “So, we’ve taken the approach of providing a software positioning engine that can be fit onto any hardware.”

Hexagon is developing products in partnership with STMicroelectronics, using the company’s Teseo V family of measurement engines. “ST is one of the established leaders of automotive GNSS solutions,” Heidinger said. “We take their measurements and put our positioning and corrections solution behind that to give positioning with lane-level accuracy.”

Noble agrees on the importance of knowing the reliability of a vehicle’s GNSS-based lane accuracy. The prevailing approach, which fuses data from GNSS and other sensors, makes it acceptable for one data source to be temporarily unavailable if the system is aware of that outage, he said. “That is where you start to see Swift, and others as well, focusing on the notion of integrity.”

An AV’s level of autonomy determines its behavior during GNSS outages. For systems with Level 2 autonomy and below, the driver must remain engaged, while Level 2+ and Level 3 systems will alert the driver to retake control when needed. If a driver of a Level 2+ or higher system fails to reengage, the AV’s reaction depends on the system and manufacturer.

“When we start to see Level 3 or above self-driving systems come onto the market, they will require that the GNSS component has an ISO 26262 safety certification,” Ruff said. “Many companies, including Trimble, are going through, or have gone through, the process of safety-certifying their offerings. As part of the AV system’s safety architecture, they will build in the capability to safely curb the vehicle if the system detects a malfunction or a spoof or some other type of problem.”


Automation Levels

In 2014, the international Society of Automotive Engineers released a standard, adopted in 2016 by the U.S. National Highway Traffic Safety Administration, that classifies cars in six levels, ranging from Level 0 (no automation) to Level 5 (full automation, meaning vehicles that can handle the full spectrum of road and traffic scenarios without any assistance from the driver). While many production models already incorporate various forms of Level 1 driver assistance, no current production car exceeds Level 2, or partial automation, which requires the driver to monitor the vehicle’s surroundings and take over as necessary. No test vehicle has yet achieved Level 5.
Image: GPS World

Image: GPS World


Other Sensors

Beyond lane-level positional accuracy, safe driving also requires avoiding collisions with other vehicles in the same lane or straying into it. Cameras, lidar and radar will detect other vehicles as well as fixed infrastructure and random obstacles, measure their distance, and monitor their movement.

While lidar scanners are better than cameras as detecting sharp-edged features, such as curbs, cameras are better at detecting and interpreting visual cues, such as road signs and the location and curvature of lane markers. In bad weather, radar is essential, because radio waves, unlike light waves, can penetrate rain, snow, fog and even dust, enabling radar to “see” where cameras and lidar cannot. However, radar sensors cannot see much detail, and cameras do not perform well in conditions with low light or glare.

Besides providing data about a vehicle’s trajectory, inertial navigation systems (INS) also measure its attitude (roll, pitch and yaw), enabling the software to better correlate and interpret data from the other sensors.

For example, when a car brakes sharply, its front end goes down; any forward-facing sensors measure distances to points closer to the car than they did a moment earlier, when its chassis was parallel to the street surface.

INS can also detect unsafe conditions, such as excessive slip angle, which is the angle between the direction of the rolling wheels and the vehicle’s true heading. A slip angle as small as 0.5 degrees can trigger skidding, spins or rollover, especially in the case of SUVs and tall trucks. Wheel-speed sensors also help verify the vehicle’s movement.

“All these technologies have their limitations,” Ruff said. “However, if you design the system, including all these technologies, then you can come up with a robust, safe combination that will enable autonomous driving.”

In addition to helping to avoid collisions, these other sensors provide relative positioning by comparing the images they acquire with highly precise maps to help locate the vehicle, especially in urban environments, which are well mapped and rich in recognizable landmarks.

Imagine an AV moving through different environments. It might travel from a city with urban canyons that degrade GNSS navigation, yet with landmarks that help relative positioning, to a rural environment devoid of both. The AVs’ algorithms must constantly weigh how much to rely on the different sensors. “Many of the OEMs and car companies are seeing that even rain mist on a highway is very bad for lidar and cameras, because it creates a big blur, but that is where GNSS will perform really well because it is open sky,” Heidinger said. “So, the two types of sensor systems complement each other very well.”

“Odometry sensors, such as a wheel-speed sensors, minimize any potential drift and add robustness to data that may have a GNSS outage of greater than 5 seconds, such as longer tunnels,” said Wesley Hulshof, principal engineer – ADAS Testing at Racelogic.

Photo: Racelogic

Photo: Racelogic

Noble sees a split in the industry. Companies such as Waymo and Cruise are pursuing Level 5 autonomy and are “heavy users of lidar” as well as other sensors. Companies such as Swift are focusing on Level 2 and Level 3 series production vehicles. “If you are making a mass manufactured vehicle for the production market, it rules out using a lidar sensor,” Noble said. “It is just too costly and complex right now to use. So, typically, if you look at the systems that are out on the market today, such as a Tesla Autopilot or a GM Super Cruise, they are very reliant on the camera as the primary sensor. Obviously, also inertial and some use of radar.”

Maps and Communications

While accurate and up-to-date maps have an important role to play in making autonomous driving possible, the more detailed maps are, the more the world they describe is constantly changing.

Meanwhile, the sensors keep improving and dropping in price, making maps less important. In the end, AVs — like human drivers — will probably rely much more on their ability to “see” and analyze their environment moment-to-moment.

Also like their human counterparts, they will gain experience. Unlike human drivers, however,  AVs will be able to instantly share their experience with every other vehicle in their area via vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communications.

V2V communications will enhance safety by informing AVs of the trajectories of nearby vehicles. If a vehicle is speeding toward an intersection and not slowing for a red light, it will be communicating its position and trajectory to other cars over a V2V link, Ruff explained.

“Then your car can make the intelligent decision to pump the brakes and avoid that collision. The same positioning stack that operates as part of the AV stack can also be used to support V2V-type applications, and the position of the vehicle will be much better than what the current V2V spec states.”

Different Approaches

Each GNSS manufacturer is taking a different approach to AV positioning.

The worlds of traditional automotive positioning and the products on which NovAtel has historically focused are coming together, Heidinger said. “The autonomous technology is demanding it and pushing for higher performance and safety-of-life functionality. Hexagon is bringing high-performance positioning solutions to the automotive industry in a manner that accepts automotive manufacturability, quality and efficiency.”

The company has also joined the 5G Automotive Association (5GAA), a large consortium developing AV solutions. “There are probably 100 companies in the industry coming together and helping to develop that vehicle-to-network communications solution, including telecom partners and automotive partners, and we are providing the GNSS expertise,” Heidinger said. “To meet the high-volume production-intent applications, including automotive quality, we recently developed a receiver based off the ST Teseo V family of measurement engines. We have an ST Teseo V set of chips on the PIM 222A product that launched in May geared exactly toward the automotive market.”

By contrast, Trimble is not focused on providing GNSS receivers or other hardware. “We allow the Tier 1 automotive manufacturers to architect the system using the components that they have selected from their preferred suppliers,” Ruff said. “We tailor our positioning solution to work with their architecture. So, we are agnostic as to the selection of the GNSS receiver, the IMU, the operating system running on the host system, and the host processor that runs the software. We can adapt our stack to run on virtually any system, using measurements from any GNSS source that meets our API requirements.”

For Swift, its “vision from day one has been to bring this type of precise positioning technology to mass market applications, such as automotive, which is a big focus for us,” Noble said. “That includes autonomy, but also ADAS, HD navigation and V2X. We do not want to be a hardware supplier in the automotive supply chain. Our boards are focused on professional and industrial markets.”

Swift’s automotive software, called Starling, runs on the vehicle’s computer. To generate a precise position, it ingests raw sensor data, as well as corrections data from the company’s Skylark network. “We focus on providing a precise-positioning stack that layers on top of any of this current generation of low-cost, automotive-grade receiver hardware from companies like STMicroelectronics.”

This test in London shows the value of inertial and wheel speed sensors. (Image: Racelogic)

This test in London shows the value of inertial and wheel speed sensors. (Image: Racelogic)

The Future

Speculation abounds as to when AVs will enter mass production and how the transition from human to robotic drivers will take place. “There might be a ‘classics only’ lane in the future,” Heidinger said “that will be the only place where cars are allowed to be driven manually.”

Safety-enhancing automotive devices typically start out as optional extras, then get incorporated into best-practice standards promoted by independent bodies. Eventually, they become compulsory.

Some automakers have committed to creating their own AVs, while others are intent on creating a turnkey solution to transform conventional cars into driverless models. However, the initial market for AVs likely will be commercial fleets rather than individual consumers.

“It will still take quite a few years before we see cars take over and drive themselves, because legislation, insurance and these sorts of things will have to happen along with the technological advances,” Heidinger said. “But the positioning side is becoming more defined. We are seeing things like L5, the Galileo constellation, coming in and becoming more available. There are more constellations providing more data for use in our solutions, so that is promising.”

Swift’s Noble said, “Most of the major manufacturers working on Level 2+ and Level 3 systems are realizing that precision GNSS will be a key component of their architecture. Most of the major OEMs have signaled some level of intent to integrate this technology. Most are tracking to start the program next year,” he added.

“We envision that in five or six years every vehicle will have a single positioning utility on board that will serve all the location-aware applications on the car — whether it is an autonomous vehicle, V2V or V2I,” Ruff said. “It will meet the most stringent accuracy requirements from all the applications and serve navigation, telematics, security, V2X and AV/ADAS applications.”

Publicerad den Lämna en kommentar

Testing autonomous vehicles inside and out

A test of Racelogic’s parking assistance system. (Photo: Racelogic)

A test of Racelogic’s parking assistance system. (Photo: Racelogic)

Racelogic helps vehicle manufacturers develop autonomous vehicle technology and test them on indoor test tracks and the open road.

Racelogic helps vehicle manufacturers develop autonomous vehicle (AV) technology and testing houses test them. Over time, regulatory and consumer testing has evolved from indoor test tracks to outdoor open-road tests, and then to indoor controlled test environments.

“Due to their application, advanced driver-assistance systems (ADAS) originated and are still mainly developed and assessed on open-sky, controlled test tracks, tackling the most common killed and seriously injured (KSI) accident types,” said Wesley Hulshof, principal engineer – ADAS Testing at Racelogic. “These assessments usually make use of sophisticated driving robots for closed loop, centimeter-accurate path following and precise speed-controlled test-track assessments. The robots can only attain this accuracy by being fed the speed and positional data by GNSS sensors, such as the Racelogic VBOX.”

Racelogic’s VBOX GNSS receiver. (Photo: Racelogic)

Racelogic’s VBOX GNSS receiver. (Photo: Racelogic)

Accuracy is key to conducting assessments for the European New Car Assessment Programme (Euro NCAP) and the U.S. National Highway Traffic Safety Administration (NHTSA). Using GNSS in conjunction with RTK base stations provides centimeter-level accuracy in position, said Hulshof, as well as accurate speed and heading information to measure ADAS data to both static and moving targets. Additionally, combining a GNSS receiver with an inertial measurement unit (IMU) allows for low-drift, high-accuracy speed and positioning information within areas of high GNSS multipath or temporary occlusions, such as gantries, bridges, forests or built-up areas.

However, “people do not just drive on closed test tracks with accurately positioned targets and infrastructure,” Hulshof said. “They do not drive at a constant throttle position and maintain an exact time-to-collision to the vehicle in front of them, like robots do. In fact, people often drive erratically.”

For these reasons, testing houses are conducting supplementary assessments on the open road, under real-world conditions. In these conditions it is still important to know vehicles’ positions and speeds to localize them and validate the system’s sensors, networks and algorithms.


Testing Stages

Stage I: Controlled
ADAS was developed for outdoor use because this is where car crashes occurred. For this, an open-sky GPS signal was essential for positioning. The types of tests and level of scientific rigor meant that the tests could be performed on closed test tracks.

Stage II: Randomized
Tests were brought to the open road to add elements not found within a closed environment such as traffic and higher speeds of the vehicle under test. For this, extra sensors were employed to add robustness in areas of obscured GNSS coverage.

Stage III: Controlled
Testing is brought back indoors for climate control and to assess L3/L4 AD functionalities such as valet parking.


Because open-road testing does not permit being constantly within range of a static base station, Racelogic developed a moving base solution for open-road testing that gives accurate relative positioning between two or more vehicles.

The increased demand for real-world testing of ADAS has generated demand for reliable ground truth data. “For example, if you consider a car driving on the winding roads of the Italian Alps and the position is out by 2m,” Hulshof said, “that is the difference between lovely scenery and falling off the side of a cliff. So, you need centimeter-level accuracy in the positional algorithms of the self-driving car, but also in the assessment tools, while we are testing it. For that reason, we still need GNSS and would ideally need RTK.”

To meet this demand, Hulshof said, Racelogic produced its own networked transport of RTCM via Internet protocol (NTRIP) solution, consisting of a modem and associated service provider. It allows for global coverage of high-accuracy, absolute positioning of a test vehicle in open-road conditions. Both the NTRIP and the moving base solutions allow ADAS testing to centimeter-level accuracy on the open road without the need to be in radio range of an RTK base station, thereby greatly expanding the testing possibilities.

“Whilst both the NTRIP and the moving base options allow for high-accuracy positioning,” Hulshof said, “they are still reliant on having an open sky for good GNSS coverage. IMU integration allows for improved accuracy over short periods of occlusion, but to truly give as accurate a signal as possible we need to be open to accept information from multiple satellite sources. That is why highest longevity accuracy is only achieved by using the GPS, GLONASS, Galileo and BeiDou constellations to provide the best RTK positioning performance in areas where that was not previously possible.”

To control the environment and allow for year-round testing, test laboratories such as the Insurance Institute for Highway Safety (IIHS) facility in Arizona and Asta Zero in Sweden have purpose-built covered test facilities, giving shelter from extreme heat or cold. Testing inside both set-ups, however, still relies greatly on the test vehicle positioning. Standard positioning techniques via GNSS in these situations is simply not possible. Therefore, Hulshof said, Racelogic designed the VBOX Indoor Positioning System (VIPS), which allows for seamless testing indoors or outdoors. “Because this system works as an alternative to satellites, with the in-vehicle VBOX allowing RTK-level performance without GNSS, the test vehicle can travel from open-sky outdoor testing to a closed environment seamlessly, with no drop in data during the transition or afterward.”

Finally, Hulshof said, ADAS and AD systems have moved on from straight-line highway scenarios to low speed turning scenarios often performed away from the open sky previously required for accurate GNSS coverage. Examples include multi-story parking garages and valet parking. “Scenarios such as self-parking and park-assist assessments, as well as indoor L1 ADAS, are becoming increasingly common requests by manufacturers on test facilities.”

These environmentally controlled facilities can simulate real-life conditions that affect specific sensors — such as sensor flare, fog, mist and water films. These types of facilities use VIPS to give outdoor GNSS accuracy in an indoor controlled environment. “There is a trend toward bringing the testing from closed test track to randomized real world back into a highly contained, climate-controlled area,” Hulshof said. “We then have an option for anything.”

Publicerad den Lämna en kommentar

Thales Alenia Space to assess feasibility of EGNSS integrity service

Image: loveguli/E+/Getty Images

Image: loveguli/E+/Getty Images

Thales Alenia Space, a joint venture between Thales (67%) and Leonardo (33%), has been selected by the European Commission for a new strategic contract to assess the feasibility of an integrity service to complement the European Global Navigation Satellite System (EGNSS) High Accuracy service, which will pave the way for use in autonomous vehicles.

Thales Alenia Space will focus on the development of a sensor-fusion approach, including and complementing evolutions of EGNSS High Accuracy. These service evolutions are aimed at providing the integrity level to serve the high-reliability and high-accuracy positioning needs of new, demanding applications such as autonomous vehicles on the road and autonomous transport in the maritime and rail sectors.

With this contract, Thales Alenia Space will assess the extension of the integrity and safety-of-life services for aviation into the road, rail and maritime sectors. In 2020, the company won the EPICURE project, based on an integrity concept for road travel (tolls and insurance), as well as the IMPRESS project, targeting an integrity service for rail signaling and train separation.

Thales Alenia Space has been a prime contractor for EGNOS (European Geostationary Navigation Overlay Service) for 25 years. It is a lead industrial contributor to the Galileo system and its ground mission segment and responsible for providing six Galileo Second Generation satellites. In April, the company was awarded a contract to support the implementation and experimentation of the navigation algorithms that will be used in the Galileo Second Generation program.

Publicerad den Lämna en kommentar

Hexagon’s new HxGN Mass Transit improves public transportation operations

System optimizes field operations and monitoring of assets through 3D, AI and mobile capabilities

Hexagon’s Safety, Infrastructure & Geospatial division has introduced HxGN Mass Transit, a geospatial transportation infrastructure management system with 3D and artificial intelligence (AI) capabilities for visualizing and analyzing transit and rail assets and operations.

HxGN Mass Transit serves as a single source of truth for infrastructure data, enabling rail-bound and transit operators to easily inspect, validate and share information on the fly.

HxGN Mass Transit combines asset and spatial data from various business systems into an integrated system, allowing operators to visualize and analyze their entire network and services. It reduces data duplication, provides access to accurate and up-to-date information and delivers greater efficiency for managing data, workflows and transit networks and operations.

Image: MarcelStrelow/iStock/Getty Images Plus/Getty Images

Zurich is using HxGN Mass Transit for its trams and buses. (Image: MarcelStrelow/iStock/Getty Images Plus/Getty Images)

Now in Zurich and Frankfurt

HxGN Mass Transit is already delivering benefits to public transportation organizations.

“Every day, we transport more than 900,000 passengers around Zurich on our 510-kilometer network with 75 tram and bus lines,” said Daniel Steger, head of electrical infrastructure, Zurich Public Transport. “Maintaining our infrastructure is vital. HxGN Mass Transit will allow us to monitor rail tracks, overhead cables and the condition of bus stops to ensure we keep the citizens and visitors of Zurich moving.”

“HxGN Mass Transit is an essential tool for managing our assets,” said Dominik Rabenau, head of data management at VGF Frankfurt’s infrastructure division. “The mobile application provides easy monitoring and the ability to update information of our timetable displays located at all stations, platforms and stops.”

Typically, transportation agencies must rely on different data sources spread across multiple systems, departments and formats. This prohibits the ability to view data in real time, making it difficult to gain a holistic view of asset conditions and coordinate maintenance.

Digital Twin of City Network

Built on top of Hexagon’s M.App Enterprise, HxGN Mass Transit overcomes these challenges. It goes beyond a simple map, providing an advanced digital twin of a city’s entire public transportation network – from track, stops and switches to construction sites, ticket machines, benches and garbage cans. It offers capabilities and workflows for supervisors, analysts, asset and operations teams and others.

“Urban population growth, increasing demand for mobility options and a greater focus on sustainability have driven interest and investment in public transportation,” said Steven Cost, president, Hexagon’s Safety, Infrastructure & Geospatial division. “By improving the ability to visualize and understand networks in real-time, HxGN Mass Transit provides a solution to the global demand for more efficient and effective public transportation.”

HxGN Mass Transit is available worldwide now.

To see a demo of HxGN Mass Transit and learn best practices for managing data, workflows and transit networks, attend the session “Driving Smart, Real-time Data Through Public Transit Systems” at the HxGN LIVE Resiliency Series, a free virtual event focused on helping critical service providers achieve greater resiliency in operations. Register for the event here.

Publicerad den Lämna en kommentar

Innovation: Attitude determination and RTK positioning using low-cost receivers

Getting It Better

Attitude Determination and RTK Positioning Using Multiple Low-Cost Receivers with Known Geometry

By Xiao Hu, Paul Thevenon and Christophe Macabiau

Innovation Insights with Richard Langley

Innovation Insights with Richard Langley

DO YOU HAVE THE BEST POSITION AND ATTITUDE? No, I’m not talking about your personal life. I’m talking about the location and orientation of a vehicle. We all know that GNSS, perhaps augmented with additional sensors, can provide the position of a vehicle accurate at the few-meter level or so under good conditions when using only single-frequency pseudorange measurements. After all, this is how most consumer-grade receivers work, including those embedded in cell phones. Yes, we can sometimes do better by using dual-frequency measurements such as those obtained by newer cellphone models with dual-frequency receivers, but we need to use carrier-phase measurements to get a significant improvement in positioning accuracy.

But this is not news. That carrier-phase measurements had the capability to provide up to centimeter-level accuracy was demonstrated in the early 1980s, when only a few prototype GPS satellites were in orbit. Surveyors and geodesists developed a relative or differential positioning technique using a reference station and one or more rover stations to determine the three-dimensional baselines connecting the stations. Knowing the coordinates of the base station, the coordinates of a rover station could be determined. This use of differential carrier-phase measurements was actually pioneered in the radio astronomy community through the development of interferometric techniques for improving the resolution of radio telescopes, which led to the invention of very long baseline (radio) interferometry (VLBI) in 1967. Geodesists realized that the astronomers’ technique could be used to obtain very precise (and accurate) baselines between radio telescopes, even if they were on different continents.

The surveyors and geodesists performing the initial baseline determinations using GPS carrier-phase observations even developed ingenious techniques for resolving the integer ambiguities that bedevil carrier-phase measurements. Those techniques evolved into the real-time kinematic or RTK positioning capability widely used today, as well as attitude determination. It was in March 2001 that we featured an Innovation article on GPS attitude determination. As we said, “[GPS] is well known for its ability to determine a platform’s position and velocity with high accuracy. Less well known is the ability of GPS also to provide the orientation of the platform. Using three or more antennas feeding separate receivers, or separate channels in a single receiver, the baseline vectors connecting the antennas can be determined. The directions of these vectors determine the platform’s three-dimensional orientation… If only two antennas are used, then only two angles or directions of the platform can be determined, such as the azimuth or heading of the platform and its elevation angle or pitch.”

While RTK positioning and attitude determination is readily done with high-end equipment, it is still a challenge to get good results for kinematic platforms with low-cost receivers. In this month’s column, we learn how a team of researchers in France is trying to do just that.


Over the past decade, GNSS has been commonly used in various domains: automotive, aviation, marine, precision agriculture, geodesy and surveying, and so on. This includes autonomous driving, which has become more and more a central topic for the automobile industry, a kind of application for which information about precise position and attitude is essential. However, the accuracy and integrity that a low-cost GNSS receiver can provide in a restricted urban or indoor environment is far from satisfactory for applications where bounded decimeter or centimeter accuracy is envisioned.

To reach this level of accuracy, techniques using raw carrier-phase measurements have been developed. Such measurements are more precise than pseudorange (code) measurements by a factor of about a hundred. Nevertheless, they are also less robust than code measurements and include a so-called integer ambiguity that requires implementation of an integer ambiguity resolution (IAR) process to use them for positioning. In some harsh environments, severe multipath and losses of lock of the receiver tracking loops create carrier cycle slips, which result in sudden changes of these ambiguities. If not detected, cycle slips create a bias in the carrier-phase measurement resulting in a reduction of position accuracy. Even if a cycle slip is detected, the IAR process has to be, at least partially, re-initialized, leading also to a loss of positioning accuracy. To increase confidence and to accelerate the IAR process by limiting the search space, restrictions can be established by using an array of two or more receivers with prior known and fixed geometry, which includes the lengths of the baseline vectors between the antennas of the receiver array and the orientation of these vectors.

In recent years, several studies have focused on the use of an array of receivers for attitude determination. However, to the authors’ knowledge, very few research articles can be found that address the use of an array of receivers to improve the accuracy of positioning or for some steps of precise position computation for real-time kinematic (RTK) processing with vehicle attitude determination, such as cycle-slip detection or integer ambiguity resolution. The objective of the research reported in this article is to explore the possibility of achieving precise positioning with a low-cost architecture: using multiple low-cost receivers with known geometry to enable vehicle attitude determination and improved RTK performance.

SYSTEM GEOMETRY AND CONFIGURATION

With the intention of performing precise attitude estimation, we have adopted a dual antenna set-up, where two GNSS antennas with a known baseline length are mounted on a vehicle’s rooftop to get an attitude estimation. Furthermore, the absolute position accuracy is augmented using the RTK approach, in which the vehicle is positioned relative to a third receiver, whose position is static and known, used as a virtual reference station (VRS). By knowing the position of the VRS, the vehicle can be positioned absolutely, too. In the positioning algorithm, we strongly rely on carrier-phase positioning which, thanks to its low noise characteristics, may enable decimeter-level positioning. FIGURE 1 shows the typical geometry of our measurement set-up.

FIGURE 1. Geometry of the model including the definition of the attitude of the vehicle.

FIGURE 1. Geometry of the model including the definition of the attitude of the vehicle.

According to Figure 1, the two GNSS antennas on the vehicle’s rooftop span the array antenna baseline b12, which one can resolve for the Euler attitude angles to get the vehicle’s orientation (heading and pitch ).
For each time epoch, we estimate both the RTK position and receiver array attitude. The baseline b13 spanned between one vehicle antenna and the VRS antenna enables us to locate the position of the vehicle relative to the VRS.

MATHEMATICAL MODELS

The realization of GNSS navigation is typically based on a Kalman filter, which is the most popular choice given its optimality and simplicity of implementation. In our study, we developed a position and attitude determination algorithm based on an extended Kalman filter (EKF).

State Transition Model. The state transition or state-space model describes how the states or parameters of the system vary over time based on a specific linear model. In our EKF modeling, the state vector includes five vehicle state parameters and 2 × (Nsat – 1) satellite state parameters: the 3D position of GNSS receiver 1 relative to GNSS receiver 3 (), the pitch angle of the vehicle (), the heading angle of the vehicle (), the DD integer ambiguities of the visible satellites seen by GNSS receiver pair 1-3, and the DD integer ambiguities of the visible satellites seen by receiver pair 2-3. Note that at a given epoch, we can consider different sets of satellites visible for the receiver pairs.

Transition Model for Position- and Attitude-Related State Parameters. In our EKF modeling for the position- and attitude-related state parameters, we suppose that they follow a random walk model, meaning that the speed and the angular rate are a zero-mean Gaussian process.

Transition Model for Satellite-Related State Parameters. The satellite-related parameters are all assumed to be constant over subsequent epochs with very small noise compared to the position- and attitude-related state parameters. The resulting state transition matrix is then given by an identity matrix and different values of process noise variance are added to complete the model.

Measurement Model. The measurement model describes how the individual sensor measurements are related to system states. In general, for every epoch, the measurement vector, which contains all measured values, can be described as a function of the state vector combined with the measurement noise vector, which describes the expected Gaussian noise of every measured value with an associated measurement noise covariance matrix.

In our model, the measurement vector comprises the following measured values: the double-difference (DD) pseudorange (code phase) measurement vector of receivers 1 and 3, the DD pseudorange measurement vector of receivers 2 and 3, the DD carrier-phase measurement vector of receivers 1 and 3, and the DD carrier-phase measurement vector of receivers 2 and 3. The DD measurements are obtained by differencing the single-difference (SD) measurements in the usual way.

In this measurement model, the position of receiver 2 is expressed in terms of the position of receiver 1 and the baseline vector between the two receivers of the array, such that it contains the known array baseline length information and the attitude information that we want to estimate. The individual DD corrected pseudorange and carrier-phase measurements for our short baseline case (less than 3 kilometers) can be modeled using certain approximations such as ignoring the tiny differential atmospheric effects.

To reflect the difference in precision between the pseudorange and carrier-phase measurements, a fixed weighting factor of 1/100 is applied to the pseudorange.

An elevation-angle-dependent measurement noise variance between all satellites is defined to complete the measurement model, defining the measurement covariance matrix.

The relationship between the state and measurement vector is obviously non-linear, thus we need to linearize this measurement function and obtain the measurement (Jacobian) matrix for use in the EKF, as usual. Our algorithm is described in more detail in the proceedings paper on which this article is based (see Further Reading).

Two alternating steps, which are the state prediction step and the state update step, are then conducted to complete the proposed EKF algorithm.

RTK PROCESSING

We first carry out a cycle-slip detection and repair scheme based on multi-epoch measurements. In our work, we use the differential phases over time cycle-slip resolution method. It is based on the observation of the differential phases between two adjacent epochs, which should include the actual jump in the ambiguity if there is one, plus some clock errors and the remaining noise term.

Except for the ambiguity term, all terms contributing to the time-differenced phases change slowly. Any cycle slips will lead to a sudden jump in the time difference of the phases. Based on the past observation of differential phase measurements, a prediction of the current differenced data can be obtained by polynomial extrapolation or interpolation. The residual between the prediction and the observation can then be used as a detector metric, to be compared to the detection threshold to decide whether there are any cycle slips.

After the cycle-slip detection process, a cycle-slip validation and size determination process is conducted to verify the determined sizes of the cycle slips. Cycle slips can be repaired using integer vector estimation similar to ambiguity resolution in the position domain.

After repairing the existing cycle slips, the RTK processing begins. From the previously described EKF process, we first obtain a float estimation of the DD integer ambiguity. The accuracy of the position state estimate is further improved by fixing the DD ambiguities to integer numbers by using the well-known LAMBDA algorithm.

The integer candidates are selected based on the sum of squared errors to get a fixed solution. The candidate with the lowest error norm is chosen once the ratio of the maximum a posteriori error norm between the second-best candidate and the best candidate is bigger than a threshold. It is a pre-defined critical value that the squared norm of ambiguity residuals of the best and second-best candidates should exceed to validate the integer estimation. In our work, we take an empirical fixed value of 3.0.

Once the IAR process is declared successful, a new position is computed using the DD carrier phase measurements corrected by the validated DD integer ambiguities. This final position is a fixed solution. If the IAR process is not declared successful, the final position is kept as the float solution.

SET-UP AND SCENARIOS

In this section, we discuss the verification of our precise position and attitude determination algorithm with real measurements from two low-cost GNSS receivers and one high-end GNSS receiver.

Data Collection. To investigate the feasibility of our proposed precise positioning and attitude determination algorithm, we set up a measurement campaign using four low-cost GNSS patch antennas and took measurements by recording single-frequency (L1) GPS pseudorange and carrier-phase measurements simultaneously at a 1-Hz rate. We put the patch antennas at various distances from each other, ranging from 60 centimeters to about 2.0 meters. In addition, we made the measurements in several different environments including an urban environment, a suburban environment, and an open-sky environment.

FIGURE 2 shows a typical set-up of the GNSS antennas for one of the measurement sessions. According to Figure 1, any two GNSS antennas on the vehicle’s rooftop span the vehicle antenna baseline, which we can resolve for the pitch and heading attitude angles to get the vehicle’s orientation. Additionally, the baseline spanned between one vehicle antenna, and the VRS antenna is able to position the vehicle relative to the latter. If the VRS antenna location is known, the absolute position of the vehicle can be determined.

FIGURE 2. Real data collection set-up: Four GNSS U-blox antennas and one NovAtel SPAN receiver antenna on the vehicle rooftop.

FIGURE 2. Real data collection set-up: Four GNSS U-blox antennas and one NovAtel SPAN receiver antenna on the vehicle rooftop.

The measurement test was performed with a vehicle on which the following hardware was mounted:

  • 4 low-cost GPS receivers with 1 Hz data rate
  • 4 L1 patch antennas mounted on the roof of the vehicle along its longitudinal axis
  • 1 high-end receiver on the roof of a building as the reference station for RTK processing
  • 1 high-end receiver tightly coupled with a tactical-grade inertial measurement unit with an antenna on the vehicle to provide its reference position and attitude.

Data Collection Scenarios. We collected three sets of data in different environments for the analyses described in this article.

Open-Sky Environment. To get an open sky and stable environment, the first measurement session took place on the football field of the École Nationale de l’Aviation Civile (ENAC) in Toulouse. As shown in FIGURE 3, the two receivers are static and their positions are fixed on the football field with favorable satellite visibility. This scenario was mainly used for validation of the algorithm implementation and to provide a reference for our multi-receiver system performance.

FIGURE 3. Receiver fixed position for dataset 1.

FIGURE 3. Receiver fixed position for dataset 1.

Suburban Environment. The second measurement session took place on the ENAC campus. The true trajectory provided by the high-end equipment and the corresponding satellite visibility during the data collection are shown in FIGURE 4.

FIGURE 4. Trajectory and corresponding satellite visibility for dataset 2.

FIGURE 4. Trajectory and corresponding satellite visibility for dataset 2.

Urban Environment. The third measurement session was performed when the vehicle was driven from ENAC to Toulouse’s city center. The whole trajectory in Google Earth and the corresponding satellite visibility during the data collection are shown in FIGURE 5.

FIGURE 5. Trajectory and corresponding satellite visibility for dataset 3.

FIGURE 5. Trajectory and corresponding satellite visibility for dataset 3.

EXPERIMENTAL RESULTS AND DISCUSSION

The datasets we collected allow us to investigate certain aspects of the simulations of our previous work. With our data collection approach, we could vary the distance between the two rover antennas, referred to as the array baseline length elsewhere in this article, as well as the type of environment. In this section, we address the following three points regarding the impact of these two aspects (array baseline and type of environment):

  • correlation of measurement error in the measurements collected by an array of receivers
  • improvement of cycle-slip detection and repair
  • improvement of positioning and attitude accuracy and ambiguity-fixed solution availability.

Experimental Results on the Correlation of Measurement Errors. As the measurements come from signals received by the closely placed antennas, it is safe to consider a certain level of correlation between these measurements. For example, the multipath error from the same satellite may be similar in the measurements recorded by the receivers connected to the two closely-mounted antennas.

By removing the corresponding geometric distance term from the DD pseudorange and carrier-phase observations, the correlation coefficient for the DD pseudorange and carrier-phase observations measured by the different antennas on the same satellite pair can be computed.

We found that the DD phase measurement errors are very correlated, and there is also a medium degree of correlation between the pseudorange measurements. We speculate that this might be due to the characteristic of the GNSS DD measurements.

To make our model closer to the real situation and thereby improve the EKF performance, we updated the observation covariance matrix in the EKF model by replacing the diagonal matrix with one including non-diagonal terms based on the correlation coefficient values we found.

Experimental Results on the Cycle-Slip Detection Performance. We carried out a number of tests on our single-frequency method, which has many limitations. Due to the imperfect detection of cycle slips, a larger ambiguity standard deviation value of 0.1 cycle was chosen to account for possible undetected cycle slips. A slight improvement in positioning results was achieved compared with a smaller value of 0.01.

Typical Positioning and Attitude Determination Performance. We found that the dual-receiver system performs better than the single-receiver situation and provides better solution availability thanks to the doubled observations redundancy.

FIGURE 6 illustrates the 3D RTK positioning estimation error for our multi-receiver method. As one can notice, the algorithm succeeded in outputting a positioning result with an accuracy of about 0.5 meters for the horizontal coordinates, which is acceptable for a harsh environment.

FIGURE 6. 3D RTK positioning estimation error illustration.

FIGURE 6. 3D RTK positioning estimation error illustration.

FIGURE 7 shows the estimation of the pitch and heading angles of the vehicle for dataset 1. One can see from the figure that the error between the estimated result and the true value is extremely small (less than 2 degrees), which can provide us with a relatively accurate vehicle posture by using our proposed method.

FIGURE 7. Illustration of vehicle attitude estimation for dataset 1.

FIGURE 7. Illustration of vehicle attitude estimation for dataset 1.

Experimental Results of the RTK Performance. We found that for our open-sky data, the dual-receiver array system provides better performance than the single-receiver RTK solution, thus demonstrating the usefulness of such an approach.

The results from the analysis of the datasets from the suburban and urban test drives gave us some useful information on the robustness of the multi-receiver RTK system in harsh environments.

As previously mentioned, the suburban data was collected when the vehicle was driven on the ENAC campus. The reference trajectory was provided with centimeter-level accuracy. The maximum standard deviation values, up to 10 centimeters, occur at around the 500-epoch mark, which corresponds to the zone having a minimum number of visible satellites. Generally, however, the environment is quite favorable with at least eight satellites in view for most of the time.

We have compared the results from the single-receiver system and our multi-receiver system in terms of the ambiguity fix success rate, the horizontal positioning (east and north directions), error statistics (mean, standard deviation, and 95% error bound), and array attitude error statistics for all three data collections. As an example, TABLE 1 gives the performance comparison between our dual-receiver system and the single-receiver system in the suburban environment and for different array baseline lengths. A better accuracy result is obtained in the dual-receiver situation.

Table 1. Performance comparison for different array baselines for data collection 2 – Suburban.

Table 1. Performance comparison for different array baselines for data collection 2 – Suburban.

Moreover, there is no huge difference in the accuracy of the positioning results between the different dual-receiver variations. However, as expected, the attitude accuracy does improve slightly as the array baseline length increases.

The high-end GNSS receiver with IMU provided a decimeter-level trajectory accuracy in the urban scenario. The number of visible satellites was much lower than for the other two environments. A clear uncertainty increase in the position solution was observed in the reference result during the trajectory section where the number of satellites was fewer than six. We also obtained satisfactory results from our urban scenario test. We can conclude that the use of an array of receivers with known geometry to improve RTK performance is feasible and effective.

CONCLUSION

In this article, we have presented a method that includes an array of receivers with known geometry to enable vehicle attitude determination and enhanced RTK performance in different environments. Taking advantage of the attitude information and the known geometry of the array of receivers, we are able to improve some of the steps of precise position computation.

We demonstrated through real data processing results that our multi-receiver RTK system is more robust to degraded satellite geometry, in terms of ambiguity fixing rate, and obtains a better position accuracy under the same conditions when compared with the single-receiver system.

ACKNOWLEDGMENTS

The research described in this article was supported by the China Scholarship Council. The article is based on the paper “Attitude Determination and RTK Performances Amelioration Using Multiple Low-Cost Receivers with Known Geometry” presented at the virtual 2021 International Technical Meeting of The Institute of Navigation, Jan. 25–28, 2021.

MANUFACTURERS

Our equipment included four U-blox (www.ublox.com) F9P GNSS receivers fed by U-blox ANN-MB series multi-band patch antennas, a Hexagon | NovAtel (www.novatel.com) ProPak6 SPAN GNSS receiver with integrated tactical-grade IMU and a NovAtel Pinwheel GPS-702-GG antenna, and a Septentrio (www.septentrio.com) AsteRx-U GNSS receiver with a Hexagon | Leica (leica-geosystems.com) AR20 choke ring antenna as the base station for RTK processing.


XIAO HU is a Ph.D. student in the Signal Processing and Navigation (SIGNAV) research group of the TELECOM laboratory at École Nationale de l’Aviation Civile (ENAC) in Toulouse, France.

PAUL THEVENON is an assistant professor at ENAC.

CHRISTOPHE MACABIAU is the head of ENAC’s TELECOM team, which includes research groups in signal processing and navigation, electromagnetics, and data communication networks.

FURTHER READING

(Coming soon)

Publicerad den Lämna en kommentar

Why radar is the future of autonomous transportation

By Steven Hong, Founder and CEO, Oculii

Steven Hong, Founder and CEO, Oculii

Steven Hong, Founder and CEO, Oculii

Radar has been around since the late 19th century, but today it is poised to revolutionize how autonomous vehicles (AVs) navigate the road. From its nautical origins as a tool to detect the location of ships in heavy fog to being a cost-effective way to prevent collisions in self-driving cars, radar has a wide range of applications.

For more than 30 years, carmakers and drivers have embedded radar in vehicles to assist with automated cruise control, automatic emergency braking, parking, and more. This effective, hardy technology plays a critical role in the driver experience today, and the same hardware will be used to help AVs navigate the road soon.

I believe that the next chapter of radar use in vehicles will be in the AV market, where software powered by artificial intelligence (AI) will use radar sensors to read a vehicle’s surroundings and get riders safely to their destination.

Radar Is a Market-Proven Hardware Solution

Radar has been around for so long, and the sensors we rely on in our vehicles every day are so reliable, that most drivers are not even aware that they have radar to thank for the assist on their perfect parallel parking job.

In this era of auto innovation and smart tech, the benefits of turning to this proven hardware solution abound:

  • Radar can perform well in poor weather conditions.
  • It is cost-effective, especially when compared to lidar and camera-based options.
  • Thanks to its low power requirements, adding radar sensors does not significantly impact a vehicle’s energy budget.
  • It is market-proven hardware that is robust and reliable in the field.

While competing technologies such as lidar are still years away from demonstrating that they can stand up to weather conditions and the toll that mileage takes on equipment, there is no question that radar sensors are up for the challenges of the road.

The flip side of this coin is that we also have the benefit of knowing the limits of traditional radar technology: It has poor spatial resolution, limited sensitivity, and a narrow field of view. However, this hardware can be greatly enhanced with the right software boost.

An Oculii sensor placed at the front corner of a vehicle. (Photo: Oculii)

An Oculii sensor placed at the front corner of a vehicle. (Photo: Oculii)

Unlocking the Potential of Radar with AI

Until recently, the best way to improve radar technology was to add more antennas until you got the resolution quality you were seeking. While this approach solves the problem of resolution, it introduces two other problems:

  1. Adding antennas exponentially increases a radar’s complexity, power consumption and size, while only improving performance linearly.
  2. In turn, this added complexity significantly increases the radar’s cost.

Consider the F-35 fighter jet, which relies on a radar system that costs more than the jet itself. While adding antennas may be a reasonable solution for military-operated airplanes, the consumer AV market would never tolerate the consequent cost increases. However, there is a way that existing automotive radars can be augmented with AI software to improve resolution, without increasing cost, size or power.

In the same way that AI software transformed what the automotive manufacturers were able to achieve with camera hardware, AI software can revolutionize how radar hardware is used for navigation in AVs.

Traditional radar sensors emit a constant, repetitive signal that delivers a reliable but low-resolution result. By using innovative AI software to emit an adaptive phase, modulated waveform that changes in real time, the resolution of traditional radar can be increased by up to 100 times. The key to transforming how we use radar hardware is all in the software.

Radar with AI

Reliable sensors with AI software can enable autonomous functions by augmenting the hardware that is already in today’s vehicles. What makes this solution so exciting is that it does not require a design overhaul: the smart sensors in question fit within existing radar packaging.

Augmenting radar hardware with AI can significantly improve performance while reducing the cost to the consumer. This formula — better performance at a lower price tag — has the potential to greatly accelerate the speed with which AVs make it safely to the consumer market and to revolutionize the automotive industry.

Rather than pushing forward with the development of costly alternatives that are prohibitively expensive for the consumer market, intelligent radar sensors can bring AVs to the road sooner and for more drivers.


Steven Hong is the founder and CEO of Oculii, a high-resolution radar company enabling the next generation of autonomous systems. Powered by AI, Oculii software increases the resolution of commodity radar hardware by up to 100 times and works in any environment.

Publicerad den Lämna en kommentar

How medium-definition maps help navigate dynamic roads

By Ethan Sorrelgreen
Chief Product Officer, Carmera

Ethan Sorrelgreen, Carmera

Ethan Sorrelgreen, Carmera

Since the early days of autonomous vehicles (AVs), maps — specifically, so-called “high-definition” maps — have played a critical role in their technology stack. Central to perception, localization and path planning, these highly detailed, highly precise maps provide vehicles a baseline understanding of the world around them, delivering key priors that form the basis of the AV’s navigational decision making.

These maps come with exacting standards: a 3D network graph, spatial accuracy within 10 centimeters, attribute support in the thousands, and so on. Additionally, with AV deployments becoming more frequent — covering broader, more complex driving domains — these requirements are growing ever more demanding.

Of particular import is the increased need for temporal accuracy — that is, a map’s ability to represent current conditions (as opposed to conditions at some point in time). Roads — especially urban roads — are highly dynamic environments. Things like construction, repaving, signal upgrades and, now, on-street dining constantly affect the flow of traffic.

For example, in a summer 2020 survey of New York, Carmera found 88 drive-lane-impacting events (out of a total of 251 road events) over 72 hours in midtown Manhattan alone.

A map’s failure to reflect such events and changes can have a major impact on an AV’s reliability (Will the autonomous-driving feature remain engaged?), motion-planning (Will the AV safely and smoothly navigate through/around the obstacle?) and/or path planning (Will the AV choose the most efficient route despite the obstacle?). Maintaining a map, however, is exponentially more complicated than creating it. Not only does the data need to be good, it also needs to be fast and cheap to produce.

The key to solving the fast and cheap legs of this classic “good-fast-cheap” trilemma is simplifying the initial problem, using what Carmera calls a medium-definition map. If an HD map is a map with high feature detail and high spatial accuracy, then an MD map is a map with high feature detail but a slightly lower spatial accuracy. It essentially atomizes the dense, complex HD world into discrete, manageable blocks, or “zones.”

An MD map of a California intersection showing road features — including control attributes — placed with zonal accuracy. (Image: Carmera)

An MD map of a California intersection showing road features — including control attributes — placed with zonal accuracy. (Image: Carmera)

These zones — each a logical section of the road network — become the new unit of fidelity. The MD map catalogs all the features in a zone — a traffic light with a left arrow that controls the left two lanes, a bike path, a solid median, etc. — but not their precise location in the real world.

This simplified map provides the ideal basis for a system of triaging change, which dramatically lowers the cost — in both time and money — of HD map updates. Indeed, it provides the foundation for Carmera’s change-as-a-service offering — a modular, on-demand feed of road events and map updates that plug into existing consumer or HD maps.

Because of its lower spatial accuracy, an MD map can be updated with consumer-grade tools — a camera and a consumer-grade GNSS, let’s say — coupled with basic consumer vision algorithms. Contrast that to an HD map, which requires either expensive equipment, like a lidar rig, or — in Carmera’s case — sophisticated algorithms that can convert visual and telemetric data into HD road graphs.

MD map maintenance, therefore, is relatively cheap, which is good news for those who want to use MD data for next-generation consumer applications, such as natural-language navigation, or to support sub-L4 levels of automated driving (both excellent MD use cases).

An MD map of the same interaction, showing road features—including control attributes—placed with zonal accuracy. (Image: Carmera)

An MD map of the same interaction, showing road features—including control attributes—placed with zonal accuracy. (Image: Carmera)

For HD updates, an additional pass is needed. Think of this as a tip-and-cue system: When a functional change in the map is detected (the tip), data from the identified zone is reprocessed using more complex algorithms to create the new HD vectors (the cue). In some cases — either because of customer requirements or because the change is superficial — a simple MD update may be sufficient. Thus, expensive computing resources are only deployed when needed.

This approach is equally effective for those using traditional lidar-based methods. There, the MD tip allows for targeted dispatching of lidar rigs, which results in significant cost-savings vis-à-vis the typical practice of sequential resurveying.
As technology evolves, so too will the role of the MD map.

Carmera sees a world where an AV’s onboard sensors will become so sophisticated that the HD maps’ utility may diminish. MD maps, however, will still provide vehicles key rules-of-the road relationships, helping optimize route planning and similar beyond-line-of-sight decision making. Employing this new standard now, therefore, not only makes driving safer today, it paves the way for the road ahead.

Screenshot: Carmera

Screenshot: Carmera