Leica Geosystems has released the Leica GS18 T, a fast GNSS RTK rover, as well as the latest versions of Leica Captivate field software and Leica Infinity office software.
Leica made the announcement at Intergeo 2017, held Sept. 26-28 in Berlin, Germany.
The announcement expands the Leica Captivate Experience. With the addition of calibration-free GNSS and various upgrades to the Captivate field software and Infinity office software, users continue the immersive experience with self-learning GNSS and engaging, intuitive software interfaces.
“In my business, speed is the name of the game,” said Manny Sangha, owner of Sangha Geomatics & Land Survey Inc. in Vanderhoof, Canada. ” With my GS16, I’ve been able to reduce time spent on every project. I can only see this getting faster and improving efficiency with the GS18. No longer having to level the pole nor calibrate the system, this is a real value for me and a game-changer in the industry.”
Calibration-free tilt compensating GNSS
According to Leica Geosystems, the GS18 T is the a calibration-free tilt compensating GNSS solution immune to magnetic disturbances.
GNSS measurements can be taken from any position on site, saving users up to 20 percent of time in the field over conventional surveying practices, because they no longer need to hold the pole vertical to level the bubble.
The GS18 T uses precise inertial measuring units (IMUs) and not a compass, so that users can measure with a tilted pole close to buildings, underneath cars and close to metallic objects.
With integrated quality assurance, the GS18 T records exactly how the pole was leveled during the measurement. The GS18 T then stores the values, ensuring measurement traceability and complete quality reporting.
Software updates
Fully supporting the GS18 T, Leica Captivate v3.0 field software and Leica Infinity v2.4 office software now offer users a more immersive means of control on site and at the desk.
Captivate now allows configuration of the GS18 T for all measuring and staking applications and the visualization of tilt compensated measurements. Measured data can be directly imported into Infinity or exported into a variety of formats suitable for CAD packages.
Within Infinity, users can visualise the measured data, including the creation of reports providing full traceability and quality assurance for themselves and their clients.
Exploring IMU specifications and correlating them to performance of a final product can be daunting, as differences between MEMS sensors are not always apparent. This article presents achievable performances in fusion technology across a range of IMUs among the best in their respective performance categories.
The number of available options in inertial navigation systems (INS) has grown substantially over the last several years. Major advances have been made not only in inertial measurement unit (IMU) technology, but also in the ability to exploit sensor information to its fullest extent. In both cases, the largest impact can be seen in the micro-electrical-mechanical systems (MEMS) sensors. MEMS sensors are typically much smaller, lower power and less expensive than traditional IMUs. The net result of these improvements is a proliferation of INS systems at much lower cost than were previously available and, therefore, greatly increased accessibility to technology that has historically seen limited deployment. Selecting the appropriate sensor and fusion solution for a particular application can be very challenging due to the large and confusing spectrum of solutions.
The IMUs will be examined in the context of new enhancements to sensor fusion algorithms such as the use of INS profiles. The concept of INS profiles applies environment specific constraints to improve performance in certain types of vehicles, or motion profiles. External sensors such as odometers and dual antenna operation can also aid the solution considerably, but will be unused in this analysis except for occasional comparisons. These external aiding sensors are extremely helpful in many cases and are available to use with a proprietary tightly coupled GNSS+INS solution called SPAN, but this paper seeks to evaluate what performance can be achieved without such aids.
Real-world test results will be examined using a selection of IMUs with the latest SPAN algorithms to illustrate what kind of performance can be achieved with different sensors in difficult conditions. Despite their major advances over the past few years, there are many challenges involved with utilizing MEMS technology to provide a robust navigation solution, particularly during limited GNSS availability or low dynamics. The measurement error characteristics of these devices have improved dramatically, but are still much larger and more difficult to estimate than traditional sensors. Advancements in SPAN sensor fusion algorithms have enabled these smaller sensors to achieve remarkable performance, especially in applications where environmental conditions allow for additional constraints to be applied.
This testing focuses on the land profile, meaning the constraints applied to a fixed-axle vehicle. The test scenarios were selected in such a way as to provide results for ideal, poor and completely denied GNSS coverage.
INS Profiles
GNSS and IMU sensors are only one part of the overall INS system performance. The sensor fusion algorithms used to exploit the available sensor data to its utmost capability are equally as important. In this regard, several improvements have been made to the SPAN INS algorithms to enhance performance under a variety of scenarios.
The largest addition to the SPAN product line is the introduction of INS profiles. That is, environment- and vehicle-specific modeling constraints can be utilized to enhance the filter performance. For example, the land profile, which will be examined in depth in this article, is intended for use with ground vehicles that cannot move laterally. The assumptions introduced for land vehicles, however, are not necessarily valid for different forms of movement, such as those experienced by a helicopter. Therefore, profiles have been implemented via command, and controlled as required by the user, allowing for maximum performance depending on the application at hand.
The land profile is analogous to what has historically been identified as dead reckoning. It is a method that uses a priori knowledge of typical land vehicle motion to help constrain the INS error growth. In other words, it makes assumptions on how land vehicles move to simplify inertial navigation from a six-degree-of-freedom system to something closer to a distance/bearing calculation. The land profile takes the concept of dead reckoning, models it as an update type into the inertial filter and adds a few additional enhancements.
Velocity Constraints / Dead Reckoning. Amongst other optimizations, the land profile enables velocity constraints based on the assumption of acceptable vehicle dynamics. This includes limiting the cross track and vertical velocities of the vehicle. Of all the enhancements, this is the one most colloquially referred to as dead reckoning.
In its simplest form, dead reckoning is the propagation of a position without any external input. In this forum, external input generally refers to GNSS satellites. Without external input, dead reckoning is inherently dependent on assumptions of velocity and heading to propagate the position. These solutions have evolved by integrating inertial and directional sensors to provide more local input and improve the solution propagation. This also is not a perfect method, however, as inertial sensors have their own errors that grow exponentially over time. The land profile velocity constraints explain the bulk of optimizations SPAN has made to enable dead-reckoning performance in extended GNSS outage conditions.
Explaining the velocity updates involves using the current INS attitude (); the vehicle attitude ( ) is estimated by applying the measured or estimated IMU body to vehicle direction cosine ( ). From this, the pitch and azimuth for the vehicle is estimated.Using the magnitude of the measured INS velocity in conjunction with the derived vehicle orientation, the vehicle velocity is computed, allowing the expected vertical velocity and cross-track to be constrained.
A velocity vector update is then applied to the inertial filter to constrain error growth. The effects of this method are expected to be most apparent in extended GNSS outage conditions when the INS solution must propagate with no external update information.
Phase Windup Attitude Updates. Some applications are inherently difficult for inertial sensors due to the fact that these systems are reliant on measuring accelerations and rotations in order to observe IMU errors. When traveling at a constant bearing and speed, separating IMU errors from measurements becomes challenging, so any application that does not provide meaningful dynamics is more demanding on inertial navigation algorithms. This type of condition commonly appears in applications such as machine control, agriculture and mining.
Gravity is a strong and fairly well known acceleration signal, so the real difficulty in this type of environment is managing the attitude, and especially azimuth, errors. Attitude parameters become difficult to observe when the system experiences insignificant rotation rates about its vertical axis.
External inputs can be used for providing input during low dynamic conditions when rotational observations are weaker. These are particularly helpful in constraining angular errors and include the same types used to assist in initial alignment: dual antenna GNSS heading, magnetometers, etc. However, as the goal of this testing is to demonstrate the achievable performance from a single antenna GNSS system, this type of external aid was specifically omitted.
Utilizing a patented technique for determining relative yaw from phase windup, the system is able to distinguish between true system rotation and unmodeled IMU errors during times of limited motion. This is a novel way to extract additional information out of existing sensors rather than adding more equipment and complexity.
The phase windup update is used to constrain azimuth error growth during low dynamic conditions that are typically not favorable to inertial navigation. However, it does require uninterrupted GNSS tracking and is therefore applicable only in GNSS benign environments. This approach is expected to show the greatest benefit in low dynamic conditions and be directly attributable to azimuth accuracy, but only in conditions where GNSS availability is relatively secure.
Equipment and Test Setup
We paired OEM-grade GNSS receiver cards with a selection of IMUs in different performance categories. Since the OEM GNSS platform is capable of tracking all GNSS constellations and frequencies, we configured each receiver to use triple frequency, quad-constellation RTK positioning. The receivers were coupled with a wideband antenna capable of tracking GPS L1/L2/L5, GLONASS L1/L2, BeiDou B1/B2 and Galileo E1/E5b signals.
Three IMUs were tested: an entry-level MEMS IMU (UUT1), a tactical-grade MEMS IMU (UUT2) and a high-performance fiber-optic gyro-based IMU (UUT3).
All GNSS receivers and IMUs were set up in a single test vehicle and collected simultaneously for all scenarios. IMUs were mounted together on a rigid frame, and all receivers ran the same firmware build that were connected to the same antenna.
The tests were conducted using a single GNSS antenna with no additional augmentation sources, such as distance measurement instrument (DMI) or wheel sensor. These are extremely helpful in aiding the solution, but as previously mentioned, this testing seeks to demonstrate the possible performance without the benefit of additional aiding sources. Dependence on aiding sources is a very important distinction when comparing such systems.
The GNSS positioning mode used was RTK via an NTRIP feed from a single base station with baselines between 5–30 kilometers. This was done to try to minimize GNSS positioning differences between the three systems. L-band correction signals were not tracked, and PPP positioning modes were not enabled.
A basic setup diagram of each system under test can be seen in Figure 1.
FIGURE 1. Equipment set-up (not to scale).
Test Scenarios
Four test scenarios will be examined using all the equipment and algorithms described above. They are: urban canyon, low dynamics, parking garage and extended GNSS outage.
The urban canyon test is designed to show the performance of the system in restricted GNSS conditions. The challenge to this scenario is to maintain a high-accuracy solution when GNSS positioning becomes intermittent or even unavailable.
The low dynamics test is intended to illustrate the benefits of the land profile, and specifically the phase windup azimuth updates in maintaining the azimuth accuracy.
The parking garage test will show the efficacy of the velocity constraint models over the different IMU classes as the extended outage provides no external information to the INS filter whatsoever. Again, no other aiding sources were used.
Urban Canyon Test. The urban canyon environment has been and remains one of the strongest arguments in favor of using GNSS/INS fusion in a navigation solution. Because urban canyons are common, densely populated and, of course, a demanding GNSS environment, they represent both an important and challenging location to provide a reliable navigation solution. Typically, they contain major signal obstructions, strong reflectors and complete blockages (depending on the city). For this reason, they provide an excellent use case for INS bridging to maintain stability of the solution.
During most urban canyon environments, it is typically rare to incur total GNSS outages of more than 30 seconds. Therefore, this scenario examines the stability of the solution in continuously degraded, but not generally absent, GNSS. In this case, the coupling technique of the inertial algorithms rather than quality of the IMU dominates achievable position accuracy.
The receiver platform is capable of tracking all GNSS constellations and frequencies. This provides a significant benefit to test scenarios, such as the urban canyon, where the amount of visible sky is significantly restricted. In this case, the more satellites that are observable, the more the tightly coupled architecture can exploit the partial GNSS information.
Though position accuracy between IMUs is less apparent in this condition, attitude results remain separated by IMU quality, which is a major consideration for some mapping applications such as those using lidar or other sensors where a distance/bearing calculation must be done for distant targets.
Test data for this scenario was collected in downtown Calgary, Canada. The trajectory (Figure 2) includes several overhead bridges for brief total outages and some very dense urban conditions.
FIGURE 2. Urban canyon test trajectory.
Table 1 shows the RMS error results of the three systems running both the default and land profiles. The first thing to notice is that the errors are differentiated by IMU category, though the differences are fairly small in the position domain thanks to the tightly coupled architecture. However, because GNSS information is partially available, the differences seen in activating the land profile are fairly modest, especially as the IMU performance rises.
TABLE 1. RTK RMS errors for urban canyon.
As the clearest benefits of the land profile are seen on the entry-level MEMS IMU (UUT1), these will be explored graphically in Figures 3 and 4. Figure 3 shows the position domain, and the RMS differences can be seen in a few cases where the default mode errors increased faster than the land profile. An example of this divergence is most obvious around the 1500-second mark of the test during periods GNSS is most heavily blocked.
FIGURE 3. UUT1 position error (std vs. land). Source: GNSS
FIGURE 4. UUT 1 attitude error (std vs. land). Source: GNSS
Low Dynamics Test. The low dynamics test is designed to emulate conditions experienced by machine control, agriculture and mining applications. In this situation, GNSS availability is generally not the limiting factor and can be used to control the low frequency position and velocity errors of the INS system. The difficulty is managing the attitude, especially azimuth, errors because attitude parameters are very hard to observe without significant rotations or accelerations (Figures 5 and 6).
FIGURE 5. Low dynamics test trajectory. Source: GNSS
FIGURE 6. Low dynamic UUT1 position errors. Source: GNSS
The low dynamics test was collected in an open-sky environment and consisted of traveling in a straight line on a rural road for roughly 2 km at an average speed of 10–15 km/h.
As this type of scenario provides little physical impetus, the azimuth and gyroscope biases are not observable. The reason for this is due to the use of the first-order differential equations to estimate the navigation system errors. Essentially, the differential equations define how the position, velocity and attitude errors change (grow) over time based on each other and the IMU errors. The observability of a particular update is tied to additional states through the off-diagonal elements of the derived transition matrix with the accelerations and rotations experienced by the system.
The overall RMS solution errors for RTK are provided in Table 2. As evident by the results presented, the position and velocity errors are clearly constrained by the continuous RTK-level GNSS position regardless of whether the land profile is enabled or not. The real differentiator in the land profile is the attitude performance due to the use of phase windup as a constraint. Moreover, the attitude improvements are certainly tied to IMU quality.
UUT1 exhibited a noticeable improvement in the attitude performance, while the higher performance IMUs did not. This is not entirely unexpected as the precision of the phase windup is lower than that of the higher grade IMUs.
Looking at the data graphically, Figure 7 shows the effect of land profile on positioning performance in this scenario. The two solutions are indistinguishable on the plot, and are all within standard RTK-level error bounds as was indicated in the RMS table.
Figure 7 shows the attitude accuracy with and without the land profile enabled. Again, the largest gains are seen on the entry-level UUT1, so this is the graphic shown below. This shows how the error peaks of the azimuth estimates are constrained. All the sharp corrections in each plot correspond to the vehicle turning around at the end of each 2-Km line and illustrates how much more powerful a rotation observation can be in azimuth accuracy overall.
FIGURE 7. UUT1 attitude error (std vs. land).
Parking Garage Test. This test was carried out at the Calgary International Airport and was selected to show the INS solution degradation during extended complete GNSS outages. The test consisted of an initialization period in open sky conditions to allow the SPAN filter time to properly converge, followed by a 500-second period within the parking garage. During the interval within the parking garage there were no GNSS measurements available.
Figure 8 provides a trajectory of the test environment. The time spent inside the parking structure is evident on the center bottom of the image.
FIGURE 8. Parking garage test trajectory.
Unlike urban canyon environments that contain partial GNSS information, this exhibits an extended period of complete GNSS outage. During this type of scenario, the IMU specifications become much more significant. IMU errors directly translate to the duration the solution can propagate before the accumulated low-frequency errors of the IMU grow to unacceptable levels. System performance during the outage degrades according to the system errors at the time of the outage and the system noise. The velocity errors increase linearly as a function of attitude and accelerometer bias errors. The attitude errors will increase linearly as a function of the unmodeled gyro bias error. The position error is a quadratic function of accelerometer bias and attitude errors.
Position results from each IMU are shown for UUT 1 in Figure 9. This plot shows the error with the land profile on and off. Without the land profile, the second-order position degradation in an unconstrained system is clearly visible.
FIGURE 9. UUT1 position error (std vs. land ).
By enabling the land profile, the filter constrains IMU errors by utilizing a velocity model for wheeled vehicles. With the constraints, the position errors are startlingly reduced for UUT1 and then progressively less impactful as the IMU quality increases in UUT2 and UUT3, respectively. This makes sense as the IMU error growth is progressively smaller in those IMUs, so the effect of mitigating them is also reduced.
Extended GNSS Outage Test. An extension of the parking garage test is to evaluate the performance in a much longer outage. Instead of 10 minutes, an outage of one hour was tested. Also, due to the extremely long GNSS outage bridging, the effects of adding a DMI sensor (odometer) will also be explored as they are able to be used as a major additional aiding source.
The most common measure of dead-reckoning performance is error over distance traveled (EDT). Due to the very long duration outages in this test, the errors will be reported in error over distance traveled to conform to the typical reporting method. This test was conducted in a mixture of highways and suburban streets with an average speed of 65 Km/h, incorporating a moderate amount of dynamics.
This effect can be seen over the duration of the entire outage as well in Figure 9. In this case, the points are the RMS error over several tests. and the light background shroud represents the one-sigma confidence as time progresses. The confidence increases over time as the overall distance traveled also increases.
FIGURE 10. Land profile EDT with and without DMI aid over 1-hour GNSS outage.
Results and Conclusions
In testing a range of IMUs in some challenging scenarios, this paper has sought to illustrate what kind of performance is achievable using each kind of system. An added complexity is looking at what effect certain inertial constraint algorithms have on this solution.
Although low-cost MEMs IMUs are continuing to greatly improve in quality and stability, the end application is still highly correlated to the overall performance of a selected INS system. For a great many applications, the MEMS devices in combination with a robust inertial filter can meet requirements and provide excellent value. However, some applications continue to require higher end sensors, and possibly post-processing to meet their needs.
The ability of SPAN to utilize partial GNSS measurements such as pseudorange, delta phase and vehicle constraints means even low-cost MEMs are capable of providing a robust solution in challenging GNSS conditions. However, this tightly coupled integration is limited in cases where GNSS is completely denied or when in low dynamic conditions.
INS profiles using velocity constraints, phase windup and robust alignment routines have been shown to provide substantial aid to the INS solution in tough conditions, such as GNSS denied or low dynamics. These improvements were shown to exhibit greater impact as the IMU sensor precision decreases. These abilities, in conjunction with the existing tightly coupled architecture of SPAN and the ever-increasing accuracy of MEMS, IMUs indicate that robust GNSS/INS solutions will continue to proliferate at lower cost targets. However, very precise applications such as mapping will continue to rely on higher quality sensors to meet strict accuracy requirements.
ACKNOWLEDGMENTS
The authors thank Trevor Condon and Patrick Casiano of NovAtel for collecting and helping to process the data presented in this article, and to Sheena Dixon for her tireless editing.
Manufacturers
NovAtel SPAN technology on the NovAtel OEM7 receiver is the testing and development platform for this research. NovAtel OEM7700 GNSS receiver cards and a NovAtel wideband Pinwheel antenna were employed. The inertial units under test were an Epson G320 (low-power, small-size MEMS IMU); Litef μIMU-IC (larger tactical-grade performance IMU still based on MEMS sensors); and a Litef ISA-100C (near navigation-grade IMU using fiber-optic gyros (FOG). Although all are excellent performers in their class and capable of providing a navigation-quality solution, the intent is to show the potential limitations that might arise due to the intended application.
RYAN DIXON is the chief engineer of the SPAN product line at NovAtel Inc., leading a highly skilled team in the development of GNSS augmentation technology. He holds a BSc. in geomatics engineering from the University of Calgary.
MICHAEL BOBYE is a principal geomatics engineer at NovAtel and has participated in a variety of research projects since joining in 1999. Bobye holds a BSC. in geomatics engineering from the University of Calgary.
SBG Systems has released a new generation of its advanced and compact inertial navigation systems. The Ekinox 2 series features new accelerometers and gyroscopes, enhancing attitude accuracy by a factor of two over the original Ekinox.
Photo: SBG
The Ekinox series is a line of tactical grade MEMS-based inertial navigation systems, first released in 2013. The latest improvements come from a complete redesign of the in-house inertial measurement unit (IMU), integrating cutting-edge gyroscopes and accelerometers.
With higher accuracy for the same form factor and price level, Ekinox 2 Series is designed for industrial-grade vehicle navigation, equipment motion compensation and data georeferencing. It provides a 0.02-degree roll and pitch, 0.05-degree heading and a centimeter-level position.
Applications for the Ekinox 2 include hydrography, mobile mapping and antenna tracking. With new accelerometers, this new generation has also significantly improved its resistance to vibration. Finally, the addition of the BeiDou constellation improves signal availability in Asia.
Compact and light-weight, the Ekinox Series has been designed to simplify installation operations. Configuration is made with an intuitive embedded web interface where all parameters can be displayed and adjusted. For example, users can choose a profile (vessel, plane, car, etc.), and the 3D view will provide a visualization of settings such as the sensor position, alignment and lever arms.
The Ekinox 2 Series is ITAR Free. The product line will be available during the second quarter of 2017.
Sensor role reversal: Lidar with its superior performance can replace GNSS in the integration solution by providing fixes for the drifting inertial measurement unit (IMU). Tests show its potential for terrain-referenced navigation due to its high accuracy, resolution, update rate and anti-jamming abilities. A novel algorithm uses scanning lidar ranging data and a reference database to calculate the navigation solution of the platform and then further fuse with the inertial navigation system (INS) output data.
Recent rapid advances in laser-based remote sensing technologies, including pulsed linear, array and flash lidar systems, have fostered the development of integrated navigation algorithms for lidar and inertial sensors. In particular, trajectory recovery based on lidar point-cloud matching can provide valuable input to the navigation filter. Lidar/INS integrated navigation systems may provide continuous and fairly accurate navigation solutions in GNSS-challenged environments, on a variety of platforms, such as unmanned ground vehicles, mobile robot navigation and autonomous driving.
In the case of airborne lidar/INS applications, the free inertial navigation solution is used to create the point clouds, which are subsequently matched to a digital terrain elevation model (DEM). The results are fed back to the platform navigation filter, providing corrections to the free navigation solution. This solution may be used to recreate the point cloud to obtain better surface data.
However, depending on the lidar data acquisition parameters, INS drift during the time between the two epochs when point clouds are acquired could be significant. Besides the shift in platform position, the drift in attitude angles could more severely impact point-cloud generation, producing a less accurate point cloud and subsequently poor matching performance.
This article describes a new lidar positioning approach, where the scale-invariant feature transform (SIFT)-based lidar positioning algorithm is used to match between the lidar measured point cloud and the reference DEM. The matching process is aided with fuzzy control: SIFT-based lidar positioning algorithm with Fuzzy logic (SLPF), where the threshold for SIFT is adaptively controlled by the fuzzy logic system.
Based on the geometric distribution and the range difference variance of the matched point clouds, fuzzy logic is applied to calculate the threshold for the SIFT algorithm to extract feature points; thus the optimal matched point cloud is extracted in several iterations. When there are enough matched points in the final output of the SLPF, the platform position is calculated by using the least squares method (LSM). Next, for trajectory estimation, when applying the SLPF algorithm, frequent lidar updates can be used to correct small cumulative errors from the INS sensor measurements. A Kalman filter fuses the results of the SLPF algorithm with the INS system.
This integrated algorithm can handle situations when there are less than three matched feature points being extracted by the SLPF algorithm, and yet they could still contribute to obtain a better navigation solution. Simulation results show that, compared to the existing algorithms, the proposed lidar/INS integrated navigation algorithm not only improves the position, speed and attitude-determination accuracy, it also makes the lidar less dependent on INS, which makes the navigation system work longer without exceeding a particular drift threshold.
LIDAR ALGORITHM
To eliminate the influence of INS error on the lidar positioning system, instead of creating a measured DEM based on INS ortho-rectification, we directly map the range data measured by lidar to the local stored DEM data. If a successfully matched feature point can be obtained, it means that we can get a point with absolute position and relative range towards the platform, which is similar to the satellite in GNSS positioning. After scanning of one area by lidar, when three or more such matched feature points, if not on a line, can be obtained, then we are able to form a full rank equation with the unknown variables of the platform position x, y and z.
However, due to the effect of affine transformation, the standardized range dataset collected by lidar is significantly different from the elevation dataset belonging to the same area. Figure 1 shows an example of the large difference between the two datasets from the same area when the pitch angle of the platform is equal to 5° and the flying height is 2,000 m. In this situation, the traditional flooding algorithm or constellation feature point matching algorithm is incapable of extracting matched feature points from such different datasets.
Figure 1. Comparison between SR and DEM data from the same area.
In response, we introduce the SIFT algorithm to the elevation map-matching procedure. Designed for image matching, the SIFT algorithm is invariant to scale, rotation and translation, and it is robust to affine transformation and three-dimensional projection transformation to a certain extent. Although SIFT is often used in image matching, each pixel from the image is a numerical point, which, in fact, has no difference with elevation data point. Before applying the SIFT, some processing on the lidar measured range data must be done.
LIDAR RANGE DATA
The scanning information of the lidar measured points are (α, β, r), where α is the angle between the laser beam and the negative Z-axis of the platform body frame, β is the angle from the laser beam to the plane of axis and Z-axis in body frame, r is the range between the laser head and the measured target, as shown in the opening figure.
Due to the terrain relief, the lidar range data are irregularly spaced. Therefore, it is necessary to interpolate the collected data. Here we apply the Natural Neighbor Interpolation method.
SIFT Algorithm, Fuzzy Control. For the lidar positioning algorithm, which is based on the absolute position and relative range of the ground-matched feature points, a point cloud with sufficient number of points of good geometric distribution is needed. In practice, however, the terrain undulation and the attitude of the airplane will affect the quality of the point cloud and the accuracy in the matching process. In addition, the selected threshold in the SIFT algorithm plays an important role on the quality of the matched point cloud.
A Monte Carlo simulation, shown in FIGURE 2, illustrates the impact of the threshold on the number of successful matched points (normalized) and mismatched rate. For obtaining better matched point clouds, we have introduced a SIFT terrain matching algorithm assisted by fuzzy control, as shown in FIGURE 3.
Figure 2. Relationship effect of threshold on the number of successful matched point (normalized) and error matched rate.Figure 3. Working principal diagram of SIFT terrain matching algorithm based on fuzzy control.
The algorithm mainly consists of two fuzzy logic controllers. Controller 1 calculates the initial threshold for the SIFT algorithm according to the gridded SR data terrain undulation degree λ, and the angle Θ between Z-axis in body-frame and Z-axis in navigation frame.
Controller 2, which is responsible to adaptively changing the threshold at each epoch, has two inputs. The first one is the Normalized Points Area (NPA), which represent the geometric condition of the matched point cloud. The other one is the Relative Range Difference Variance, which indicates if a mismatch has happened. When the final matched feature point cloud is obtained, and the number of points is greater than or equal to 3, then the LSM is used to calculate the position of the platform.
INS/LIDAR NAVIGATION
Loosely and tightly coupled integration are the most common methods in navigation systems. Given the characteristics of the proposed positioning algorithm, the classical integrated navigation algorithm needs to be modified. In the loosely coupled approach, the lidar is unable to aid INS when flying through a flat region and/or flying with a large tilt angle, because the proposed lidar positioning method may have difficulty in extracting enough matched points to calculate a position.
In the tightly coupled method, as the output frequency of matched point cloud is low and the geometry of the matched feature points is relatively poor, the integrated system may be extremely unstable. Here we propose a combined loosely and tightly (CLT) integrated navigation algorithm that when the lidar positioning algorithm can extract enough matched points for a navigation solution, the lidar-calculated navigation solution is used as the main observation.
However, when the matched points are not sufficient to obtain a navigation solution, the baseline vector of the matched point that is closer to the projection of the platform center to the surface will be utilized as the observation. In this solution, lidar can still provide a certain degree of aid to the INS, once extracting matched feature points, even if less than 3.
SIMULATION ANALYSIS
In the simulation experiment, the 3D DEM data of 0.5-meter resolution is obtained from an open source named EOWEB. Then the DEM data is resampled to a higher resolution of 0.1 meter, which is used to generate the simulated, irregularly spaced, measured range data. On the basis of the original DEM (0.5 meter resolution), the proposed lidar positioning algorithm and lidar/INS integrated navigation algorithm are verified and compared with the traditional methods.
Simulation of Lidar Algorithm. As shown above, the successfully matched points rate is very important for positioning, as once a mismatched point occurs, it may lead to a faulty navigation solution. In the simulation, the proposed SLPF is simulated under the condition of different aircraft tilt angle ϴ, from 0° to 10° with a step of 1° , at 5,000 different positions, which is the same simulation condition as in Figure 2. Comparison is made with the traditional constellation feature matching based lidar positioning algorithm (CLP) and the SIFT based lidar positioning algorithm without fuzzy control (SLP). The successfully matched points rate and the NPA value are shown in Figure 4.
Figure 4. Successful points matched rate and the NPA value results under different aircraft attitude condition from three different algorithms.
As can be seen from the figure, along with the increasing platform attitude angle, the successfully matched points rate of all the three algorithms has declined. However, compared to the CLP, both SIFT-based algorithms have a higher success matching rate due to the more stringent feature-point extraction approach. And due to the adjustable threshold mechanism, the SLPF could remove some of the mismatched points by raising the threshold; thus it is superior to the common SIFT algorithm in performance. The NPA values of the extracted point cloud from the three algorithms are shown in Figure 4(b). With the increased attitude angle, the NPA value of the matching feature point cloud decreases in all three algorithms. The CLP algorithm, however, is more sensitive to the projected range data, which makes the number of successful matching points drop sharply, and further affect geometric distribution of the point cloud. The gap between the SLPF and SLP shows that the fuzzy control module can help improve the geometric structure of the feature point cloud.
Figure 5 shows the positioning error when applying the three different matching algorithms at 5,000 different areas. The SLPF algorithm is better than the other two algorithms in all directions. When the platform’s attitude angle reaches about 10 degrees, the north and east positioning accuracy of SLPF algorithm is still about 8 meters, and the height positioning accuracy is about 0.2 meters. The reason that the height positioning error is far less than the north and east positioning error is because of the matching point cloud distribution. Due to the airborne lidar scanning mechanism, the matched point cloud is all located in a relative small area at the bottom of the platform, resulting in the great component value in the height direction of each matched feature point baseline vector in the G matrix, and then affect the final positioning accuracy.
Figure 5. Positioning accuracy under different aircraft attitude conditions with different algorithms.
Table 1 shows some detailed information as average number of matched points (ANMP) and matched points position error (MPPE) using the three methods. The MPPE is calculated in 3D space. It can be seen that when the tilt attitude is small, comparing to the CLP method, although the number of matched points extracted by SLPF is less, the matched points position accuracy is still much better, leading to a better localization result. Moreover, with the increasing platform tilt attitude, CLP and SLP have more difficulty in maintaining the number and accuracy of the matched points.
Lidar/INS Algorithm. To validate the feasibility of the proposed integrated navigation algorithm, firstly, the motion trajectory of the platform must be simulated. As shown in Figure 6, the red line is the simulated platform true trajectory, which lasts for 1,400 seconds. During the trajectory, the platform undertakes the different motion states as acceleration, deceleration, climbing, turning and descent. Then the INS output data based on the true trajectory with the frequency of 100 Hz is generated. To verify the calibration performance on the INS in the integrated navigation algorithm, accelerometer and gyroscope drift noise is added to the INS output data. The green line shown in Figure 6 is the INS output data trajectory solution. At the end of simulation, the error to the east direction reaches 500 meters, and the north direction error reaches to more than 2,200 meters.
Figure 6. Comparison between True trajectory and INS calculated trajectory.
At the same time of the INS outputting navigation solution, lidar also scans and calculates the position of the platform with 1-Hz frequency. Note that the speed of the aircraft is from 70 m/s to 100 m/s, and the maximum lidar scanning angle αmax is 20°. Figure 7 and Figure 8 show the number of matched points and the positioning error for each scanned terrain using SLFP. When the platform maintains smooth flying, the number of matched points can reach an average of 10, and the positioning accuracy is relatively high, less than 3 meters. Note, during the period, only in a few epochs are the number of matched points less than five. However, when the platform is climbing or changing flight direction, the number of matched points is obviously decreased due to the large tilt angle of the platform, and so does the number of successful positioning times. In this case, the position error is also increased dramatically, reaching about 10 meters error in east and north, and 0.2 meters error in height. Especially in the course of changing the direction of the flight, shown in Figure 7, during the periods of 720s–800s and 920s–1,000s, due to the larger roll angle, the SLPF could hardly be able to calculate the position through the LSM. During this period the lidar would occasionally output 1 or 2 matched feature points.
Figure 7. The number of the matched points of each lidar positioning epoch.
Figure 8. The positioning accuracy of each lidar positioning epoch.
During the simulation, the CLT and LC methods are used for data fusion and trajectory estimation comparisons. TC method is not added to the comparison because of slow convergence. The data fusion results are shown in Figure 9. It illustrates that the LC method and the CLT method have close positioning accuracy in the case of sufficient matched feature points. As can be seen in conjunction with Figure 8, when lacking matched points, the CLT method is superior to LC on positioning accuracy, especially in the height direction. In addition, the CLT integrated algorithm shows some improvement on the accuracy of estimating speed and attitude.
Figure 9a. Data fusion results using two different integrated algorithms: position determination error.
Figure 9b. Data fusion results using two different integrated algorithms:velocity determination error.
Figure 9c. Data fusion results using two different integrated algorithms: attitude determination error.
Figure 10 shows the position error distribution when using four different lidar/INS integrated navigation methods for data fusion under the condition of different simulation trajectories. In the simulation, 50 1,400-second-long different trajectories, with flat areas, are generated with different platform attitude, velocity or acceleration. As can be seen from the figure, compared to other integrated navigation methods, the CLT method greatly improves the accuracy of navigation.
Figure 10. Position error distribution when using four different lidar/INS integrated navigation method.
During 84.26% of the simulation period, CLT could maintain the position error less than 3 meters; the rate with error that is larger than 15 meters is 1.2%. For the TC method, due to the frequent divergence of the data fusion filter, most of the position estimates are not available. In addition, after flying above a flat area, the voting-based constellation integrated method has poor matched point accuracy and successfully matched rate due to large INS drift error, which makes lidar unable to calibrate the INS. When using the constellation-based method, during only 32.35% of the simulation period, the error is maintained in 3 meters and most of the period, 54.9%, the position error is between 3 to 15 meters.
CONCLUSION
We propose a new lidar matching algorithm based on SIFT, which does not rely on the INS output data to generate measured DEM data, and can adaptively change the threshold of the SIFT algorithm to generate optimal matching between the point cloud and the DEM. Through verification of simulation, the algorithm is compared with traditional lidar/INS integrated navigation methods based on comparing achieved accuracies in estimating position, speed and attitude. Simulation results show that the SLPF algorithm has better reliability for feature points matching and robustness against the platform attitude than the traditional algorithms. The CLT method improves trajectory estimation accuracy, especially when flying over moderately undulating terrain or flying with large roll or pitch angles.
ACKNOWLEDGMENT
This article is based on a paper presented at the ION International Technical Meeting, January 2017. This research used an open-source GNSS/INS simulator based on Matlab, developed by Gongmin Yan of Northwestern Polytechnical University, China.
Haowei Xu is a Ph.D. student at Northwestern Polytechnical University, where he received an M.Sc in Information and Communication Engineering. He is a visiting scholar at The Ohio State University.
Baowang Lian is a professor at Northwestern Polytechnical University where he is also director of the Texas Instruments DSPs Laboratory.
Charles K. Toth is a senior research scientist at the Ohio State University Center for Mapping. He received a Ph.D. in electrical engineering and geo-information sciences from the Technical University of Budapest, Hungary.
Dorota A. Brzezinska is a professor in geodetic science, and director of the Satellite Positioning and Inertial Navigation (SPIN) Laboratory at The Ohio State University.
I have mixed emotions as I write this column. Delighted, absolutely, to be given the opportunity to write for GPS World on topics that I am so passionate about; but also sad that we will not see any more articles from Don Jewell, whose excellent columns I followed so religiously over the years. I never had the opportunity to meet Don personally but, to me, he is irreplaceable. But let’s talk about the changing face of defense positioning, navigation and timing (PNT) — not in the editorial sense, but in the technology sense.
As we all know, PNT and GPS are no longer synonymous. With a host of innovative technologies on the horizon, PNT is about so much more than GPS these days, and the military knows it. Sure, GPS has been the workhorse of PNT for many years, and it’s not going anywhere anytime soon. I’ll be clear on that: GPS is not going anywhere. But it’s not a complete solution either.
Let me paraphrase what a friend in the infantry tells me, by saying GPS is a 60 percent solution to their navigation needs. What does that mean? Well, it goes something like this:
60 percent of the time: GPS is great, it does what we need.
20 percent of the time: We are indoors or underground, and GPS is simply not available.
15 percent of the time: We’re in an urban canyon. GPS availability is intermittent, and the accuracy is poor.
4 percent of the time: We’re in forests or dense vegetation, and GPS is sporadic.
1 percent of the time: GPS is jammed.
You can argue the numbers depending on the mission, but you get the idea. What, then, is the answer for the soldier? Well, first things first: We don’t want to reinvent the good 60 percent so, once again, GPS is here to stay. The question is how do we push past that 60 percent figure and get ourselves closer to 100 percent? Let’s go from the bottom up, and address GPS jamming.
Overcoming interference
The classic solution to jamming is an adaptive antenna, also known as a controlled radiation pattern antenna (CRPA). More on this another time but, for now, suffice it to say that CRPAs are a well-understood and mature technology, and can offer very high levels of jamming resistance.
The often-cited disadvantage of a CRPA antenna is its size, weight and power: As CRPAs employ multiple antenna elements, they are inherently larger and heavier. The electronics can pretty much be covered by a single chip these days, leaving the antennas themselves as the problematic aspect, but advances in antenna technology have also made big hurdles.
For airborne platforms, conformal antennas designed as part of the structure or fuselage can be used; whilst for the dismounted soldier, the trend is towards wearables, where the antennas may be an inherent part of the clothing or helmet design.
Aside from adaptive antennas there are a whole host of other techniques in your anti-jam kit bag, including receiver-based techniques.
It’s a numbers game
For forests and urban canyons, this is where multi-frequency multi-GNSS comes into its own. It really is a numbers game: The more constellations you use, the more satellites you can choose from, and the greater your chances of seeing enough satellites to derive a reasonable navigation solution. You also have more options for mitigating the effects of multipath and other errors.
Of course, this gives rise to a potentially difficult question for some governments: In defense applications, do you want to rely on foreign GNSS constellations as part of your PNT solution? The attitude here depends on your own country’s policy and a trade-off of perceived gains against perceived threats. The UK, for example, has chosen to embrace all available constellations and frequencies in future military navigation systems.
That’s probably about as far as GNSS gets you, because now we’re looking at the 20 percent of the time where the user is indoors or underground. In other words, environments where GNSS simply isn’t available. This 20 percent is perhaps more tricky to address, and is the realm of alternative and complementary PNT technologies.
Beyond GNSS
Fusing different sensor modalities to create a combined navigation solution is anything but a new idea. The benefits of combining GPS with an inertial sensor were recognized a long time ago, and this classic pairing continues to be the subject of research today.
The two technologies are highly complementary in various ways: GNSS offers absolute position, low short-term accuracy, and high long-term accuracy. On the other hand, an inertial sensor offers the opposite: relative position, high short-term accuracy, and low long-term accuracy. It’s a match made in heaven.
But whilst GNSS plus inertial may be a good choice for, say, airborne platforms, it doesn’t solve the in-building and underground problem. Without GNSS, you need something else.
Indoor navigation has been one of the hottest research topics of recent times, but there are really two types of indoor scenario: the first is when you’re in a shopping mall or airport. You can use an inertial sensor, Wi-Fi, mobile base stations, and various other bits of infrastructure to help you navigate.
The second scenario is the military one: You’re in an unfamiliar enemy compound or underground tunnel complex. In this case, there is no GNSS, no Wi-Fi, no mobile communications; and, for navigation, you can only really rely on the sensors you bring with you.
So what other sensor works underground, and complements inertial?
Visual/inertial integration
Visual odometry is an established, yet often overlooked, navigation technology that is undergoing a resurgence of interest, in both military and civilian applications. In simple terms, visual odometry uses sequential camera images to determine motion in a six degrees of freedom reference frame. Using either single or multiple cameras a platform can estimate both its 3D position and orientation, providing much the same information as an inertial sensor — but with a few added benefits.
Visual/inertial sensing allows 3D reconstruction of a road incident. (Screenshot: Roke)
Because cameras and associated vision-processing algorithms are capable of detecting corners and features, a 3D model of the environment in which the soldier is operating can also be built up. In other words, we can perform simultaneous localization and mapping (SLAM).
But like any navigation technology, visual odometry has its limitations. It likes well-defined features in the environment, such as corners, but can get confused by moving objects like trees and clouds. Its performance also depends on factors such as the quality of the camera and lens, and how well the system is calibrated. Like an inertial sensor, it provides a relative positioning solution and is subject to accumulation of errors over time. It’s a great technique, but it really comes into its own when combined with another navigation sensor, such as an inertial unit.
And it’s not just the military guys who are taking advantage of visual/inertial integration. Just take a look at Google’s Tango project, or what Qualcomm is doing, or Roke’s black box for driverless cars, to name but a few examples.
Bringing it all together
Over the course of the last decade or two, the operational landscape for soldiers has changed significantly, with far greater focus on urban warfare. The military realized some years ago that the answer to robust navigation for dismounted soldiers was going to require a range of sensor modalities: no single navigation technology is ideal in all environments. That’s why this has been the focus of so many defense programs of recent years.
By way of example, the UK Ministry of Defence (MoD) initiated a research program in 2013 called Dismounted Close Combat Sensors (DCCS). The contract addressed a range of soldier capabilities, one of which was the ability to provide reliable soldier position and orientation in all environments.
The DCCS programme evaluated a whole bunch of technologies, but eventually converged to an integration of three primary sensors: multi-constellation GNSS, a low-cost inertial measurement unit (IMU) and a video camera. The single monocular video camera was used to strap down the IMU, in a very tightly-coupled system. It makes sense: when GNSS is available, use it. When GNSS isn’t available, the integrated visual/inertial navigation sensor continues to provide both location and orientation for the duration of the mission. As it should be for a tightly integrated navigation system, the performance of the combined system outperforms any individual sensor in isolation.
Whilst integrated sensor systems enable our soldiers to position, orientate and navigate themselves, the performance of individual sensors continues to be pushed to new limits. Inertial technology is advancing all the time, and defense is again pushing the boundaries. Take a look at what DARPA is up to, as an example.
The missing ‘T’
Haven’t we missed something? Ah yes, there’s a “T” in PNT. So whilst there would seem to be various options for achieving a robust positioning and navigation solution, we mustn’t forget precise timing for those applications that need it. Quantum technology is flavor of the month here and, once more, the defense agencies are furthering developments: DARPA with its ACES program, and MOD/DSTL via the Quantum Technology Program, to illustrate just a couple of examples.
So whilst GPS will continue to remain the workhorse, defense PNT is migrating from GPS-only to being a many-faced beast. And I haven’t even gotten started on pseudolites, signals of opportunity, eLoran, and cooperative navigation.
The future of defense PNT looks pretty good to me.
NovAtel Inc. is showcasing its high precision positioning technology as part of AutonomouStuff’s “Roadmap to Autonomy” exhibit at the 2017 Consumer Electronics Show (CES), Jan. 3-8 in Las Vegas. The exhibit is located at the MGM Grand in the Skyline Terrace Suite.
AutonomouStuff provides research and development platforms for the safe and reliable testing of automation technologies.
It uses NovAtel’s exceptionally robust SPAN GNSS + Inertial (INS) technology to provide the highly precise, continuous 3D positioning necessary to evaluate robotic and autonomous solutions for autonomous applications.
NovAtel’s SPAN technology combines a high-performance Global Navigation Satellite System (GNSS) receiver with an Inertial Measurement Unit (IMU) to deliver deeply-coupled centimeter-level positioning. SPAN provides robustness against short GNSS outages, using IMU updates to bridge the positioning solution. SPAN also provides high data rate position, velocity and attitude (pitch, roll, heading) updates to capture the full real-time motion profile of a vehicle. Widely deployed in the automotive R&D space, SPAN supports applications ranging from autonomous navigation to V2X systems, where it is utilized to provide a source of vehicle ground truth.
As a committed technology partner, NovAtel has worked closely with AutonomouStuff to optimize SPAN for AutonomouStuff’s vehicle perception kits. As a result of these efforts, AutonomouStuff is able to offer three different levels of positioning performance — “good, better, best” — based on the grade of IMU selected.
“We are always excited to work with the team at NovAtel and cannot wait to show off their ‘good, better, best’ SPAN GNSS options for autonomy in our suite at CES,” said AutonomouStuff CEO Bobby Hambrick. “Their solutions are a significant piece of autonomous research and development. With three kit options, there is something for everybody. We’ve done the work for you, allowing you to choose which kit is best for you based on your accuracy needs and price range.”
The collaboration with AutonomouStuff is reflective of NovAtel’s commitment to the development of fully autonomous vehicles for a wide range of industries. In May 2016, NovAtel announced the formation of a new Safety Critical Systems (SCS) Group, tasked with developing functionally safe GNSS positioning products that will meet the exceptional performance and safety requirements of autonomous vehicles.
“Our team made significant progress in 2016 towards product definition, GNSS integrity for automotive applications, and corporate TS 16949 compliance,” said Jonathan Auld, Director of the SCS Group at NovAtel. “As the world leader in high precision GNSS technology for more than 20 years, NovAtel is leveraging its extensive experience developing safety critical systems for the aviation industry to meet the future safety thresholds required for driverless cars.”
AutonomouStuff and NovAtel representatives will be available in the MGM Grand Skyline Suite during the CES to answer customer questions. To set up a meeting with the NovAtel SCS team at CES 2017, attendees can contact Allan MacAulay, Business Development Manager, SCS ([email protected]).
A number of organizations are focusing on how inertial can help GNSS receivers to provide more stable, reliable position outputs when signals are hard to receive. Papers presented in September at the ION GNSS+ 2016 conference in Portland, Oregon, demonstrate that there is indeed a lot of focused effort in this area.
The conference showcased several integrated inertial GNSS solutions from a range of companies. For example, NovAtel is developing a novel way to make better use of lower precision MEMS inertial for certain land applications. Qualcomm is moving forward with a low-cost visual inertial to advance autonomous vehicle developments. And researchers in Germany from a university spin-off company are studying a multi-sensor solution.
Inertial integration aiding
Many people have heard about the NovAtel SPAN inertial/GNSS system. SPAN inertial-integration-aiding software has now been available integrated on NovAtel GNSS engines for a number of years. Combined with various external inertial packages providing real-time inertial aiding data, this system enables positioning outputs over a wider range of more difficult signal environments where GNSS alone might be too stressed to perform well.
According to the website, NovAtel currently offers SPAN with MEMS inertial products including various models from Honeywell, Litef, Analog Devices and Sensonor, along with a number of fiber-optic and high-precision tactical grade inertial measurement units (IMUs).
Recent SPAN development efforts have been focused on improving the performance of combined GNSS/SPAN/MEMS IMUs. The premise of the work is that in land-vehicle applications, a “land profile” can be applied that constrains velocity based on a range of acceptable vehicle dynamics. This includes applying limits to the cross track and vertical velocities of the vehicle.
In testing this land model, with equipment mounted in the NovAtel test van, three types on IMU were run through three different test scenarios. The IMUs were:
Epson G320 — Low power, small size MEMS IMU
Litef μIMU-IC — Larger tactical-grade performance IMU still based on MEMS sensors
Litef ISA-100C — Near-navigation-grade IMU using fiber optic gyros (FOG).
The three test scenarios involved environments with clear sky, partially obstructed sky view (downtown urban canyon) and a parking garage with no view of the sky and no satellite signal reception.
The Epson MEMS IMU appeared to be at a disadvantage from the beginning, given the higher performance units to which it was being compared. But NovAtel’s objective was to demonstrate that even this lower end device, when combined with GNSS, SPAN and the land profile, enables pretty good positioning results.
Test set-up.
Impact of SPAN land profile in GNSS denied testing.
The tests indicated that positioning with integrated higher performance units did not benefit to the same extent as when coupled with the low-end MEMS units in land-profile mode. Acceptable positioning was indeed possible with the Epson MEMS and when the constraints of land profile were able to limit position excursions when GNSS was lost, as in the parkade tests at Calgary airport shown in the figure above.
Ryan Dixon and Michael Bobye from NovAtel Inc. wrote this ION GNSS+ paper, “Performance Differentiation in a Tightly Coupled GNSS/INS Solution.” Ryan Dixon is the chief engineer of the NovAtel Synchronized Position Attitude Navigation (SPAN) GNSS/INS products, and Mike Bobye is a principal geomatics engineer at NovAtel Inc.
Visual inertial odometry
Qualcomm also presented some interesting results for the integration of visual inertial odometry (VIO) with GNSS. VIO measurements are constructed from a stream of camera frames combined with inertial measurements and can provide high-accuracy relative positioning. In experiments in a not-too-severe urban-canyon environment, this approach has been seen to reduce 95 percent horizontal error by two-thirds compared to GPS alone.
For applications such as autonomous vehicles and advanced driver assistance systems (ADAS), 50-meter errors, which can be typical for stand-alone GPS in urban canyons, just won’t cut the mustard. So Qualcomm has been looking for another source of aiding that would help reduce errors significantly.
The test set-up used a Sony Xperia Z3 phone as the source for the camera data and separate VIO processing, along with a single-frequency CSR SiRFstarIV GPS module on a custom hardware board for raw pseudorange and Doppler range-rate measurements. A high-precision NovAtel OEM6 GNSS/IMU SPAN-CPT module was used as ground-truth for position measurements.
Two scenarios were used to evaluate the proposed approach. The first scenario is an 870-meter drive in downtown Somerville, New Jersey, with a duration of 261 epochs. This represents a mild urban-canyon environment with loss of signal errors of a few tens of meters.
(Left) Part of the trajectory for the drive testing; (right) walkthrough building with no GPS coverage.
Results from the drive testing include several large GPS errors that the GPS+VIO solution is able to significantly reduce, while the walkthrough building tests appear to demonstrate a continuous GPS+VIO position solution.
“Robust Positioning from Visual-Inertial and GPS Measurements” was written by Urs Niesen, Venkatesan N. Ekambaram, Jubin Jose, Lionel Garin, and Xinzhou Wu, all of Qualcomm Research.
Multiple sensors
Finally, researchers at the Technical University of Munich (TUM) in Germany have focused on bringing outputs from as many sensors as economically feasible into an integrated GNSS solution. A precise model for multipath is included that applies amplitude, code delay, phase shift and Doppler shift for each reflected signal. The magnetometer measurements provide rough attitude information, which enables robust GNSS attitude ambiguity fixing.
The ANavS module integrates a multi-constellation u-blox GNSS receiver with a Sensonor 3D accelerometer/gyroscope/magnetometer, a Bosch barometer/thermometer and a built-in dual-band Taoglas GPS/GLONASS antenna. Real-time kinematic (RTK) positioning was tested by TUM students using the measurements from the multi-sensor module and a virtual reference station (VRS). A second multi-sensor module placed on the rear of the vehicle enabled attitude determination.
“Reliable RTK Positioning with Tight Coupling of 6 Low-Cost Sensors” was authored by Patrick Henkel, Technische Universität München, and Houcem Hentati, Advanced Navigation Solutions, Munich, Germany.
All of these options are providing GNSS with the support it needs in tight signal situations.
By Rui Sun and Hongyang Bai, Nanjing University of Aeronautics and Astronautics, and Ke Han, Jun Hu and Washington Y. Ochieng, Imperial College London. Presented at ION GNSS+ 2016.
An Integrated Algorithm Based on BeiDou/GPS/IMU and its Application for Anomalous Driving Detection
This paper introduces an integrated algorithm for detecting lane-level anomalous driving. Lane-level high accuracy vehicle positioning is achieved by fusing GPS and Beidou feeds with Inertial Measurement Unit (IMU) using Unscented Particle Filter (UPF). Anomalous driving detection is achieved based on the application of a newly designed Fuzzy Inference System. Computer simulation and real-world field test demonstrate the advantage of the proposed approach over existing ones from previous studies.
By Sashidharan Manickam and Kyle O’Keefe PLAN Group, Department of Geomatics Engineering, University of Calgary
This paper analyzes the GNSS signal authentication limits in using different grades of IMU (Tactical and MEMS) to detect errors in combination with different grades of GNSS receiver (Geodetic grade and Automotive). To test these combinations, a tightly-coupled 23 state navigation Kalman Filter is implemented with a constant velocity dynamics model for the position, velocity, attitude and clock states and first-order Gauss-Markov processes to model the 12 sensor errors.
NovAtel debuted two new inertial measurement unit (IMU) products within its SPAN technology portfolio at ION GNSS+ 2016, which was held Sep. 12-16 in Portland, Oregon.
SPAN couples NovAtel’s GNSS precise positioning technology with robust inertial navigation systems (INS) to provide continuous 3D position, velocity and attitude solutions, the company says in a news release.
IMU-µIMU-IC
The compact IMU-µIMU-IC is a high performing, fully commercial MEMS IMU. Small in size, it is suitable for aerial and hydrographic survey and space constrained industrial applications. The µIMU is available as a complete assembly in an environmentally sealed enclosure or as a standalone OEM product, both compatible with the company’s OEM6 and OEM7 SPAN receivers.
NovAtel also developed an enclosure for its Honeywell HG1900 IMU, which was previously available only as an OEM product. The IMU-HG1900 IMU offers a hybrid package of Honeywell’s micro electromechanical systems (MEMS) gyros and RBA accelerometers. The enclosure provides system integrators with design versatility, offering LED indicators and simplified cabling that can be extended in length as required. Both cabling and connectors are available off-the-shelf, NovAtel says.
“These two IMUs are part of our new IMU enclosure family, which now provides four sizes of enclosures – from the small Litef- µIMU to our high performance IMU-ISA-100C,” says Neil Gerein, portfolio manager for NovAtel. “We’ve worked hard to bring our customers the very latest in IMU technology and to expand IMU choices to ensure the optimal positioning performance for their application.”
Shipments of the new IMU enclosures will be available in Q4 of 2016, according to NovAtel.
Xsens has launched BASE, an online technology platform with a community forum and a knowledge base on 3D motion tracking technology and products.
On BASE.xsens.com, the knowledge base contains inside information about micro-electro-mechanical system (MEMS) sensors, inertial measurement units (IMU), sensor fusion algorithms, body-motion tracking and motion capture.
It also provides best practices, tips and tricks for the use of Xsens’ successful products the MTi series, the MTw and the MVN wearable motion capture solutions. A second section of BASE is the community forum with direct access to Xsens’ engineers and other Xsens users.
The knowledge base and community forum make it easier to integrate the MTi or MTw and to get the most out of MVN. If a question is not answered in the knowledge base, it is straightforward to ask a question to the community. With short response times from either other Xsens users or the entire Xsens engineering team, the user community is a quick way to continue development, Xsens said.
BASE is a next step by Xsens to support the growing community and interest in inertial technology. It further enhances the interaction between users and Xsens.
“Although Xsens makes it easy to use inertial technology in their applications, the underlying technology is complex and there are many features for specific applications,” said Remco Sikkema, Xsens marketing manager. “Understanding the technology makes it easier to integrate the products and be successful with Xsens.”
With BASE, engineers and engineering teams in the Xsens community can come closer together. The primary goal is to make Xsens customers more successful by providing a platform to exchange information.
There is no need to register for BASE to access the community forum and the knowledge base. To ask questions or comment on articles, registration is possible via SSO or email.
Micro-electromechanical system (MEMS) gyroscopes have advantages for orientation sensing and navigation as they are small, low cost and consume little power. However, the significant noise at low frequencies produces large orientation errors as a function of time. Controlled physical rotation of the gyroscope can remove the constant part of the gyro errors and reduce low-frequency noise. As adding motors for this would increase the system cost, it would be advantageous to attach gyros to a rotating platform that is already built in the vehicle. The authors present theory and results for novel navigation systems where an inertial measurement unit (IMU) is attached to the wheel of a ground vehicle. The results show that a low-cost MEMS IMU can provide a very accurate navigation solution using this placement option. It has two clear advantages:
Wheel motion removes the constant bias of the gyroscopes
Distance traveled can be estimated from accelerometer data.
For low-dynamic ground vehicles, this approach is superior to conventional dead-reckoning with an odometer when a low-cost MEMS gyro provides the heading information. Test results are obtained using a vehicle driving slowly on a relatively smooth surface, and the use of an accelerometer for wheel phase-angle tracking was fairly accurate for this purpose.
For higher vehicle dynamics and gravel roads, the accelerometer data will be contaminated with significant centripetal and motion-caused accelerations. For that purpose, the use of high-range gyro with the sensitivity axis perpendicular to the wheel plane should be considered to complement the accelerometer-based (bias-free) observations. Applying this method to passenger cars at highway speeds would require an IMU with wide bandwidth, and solving the challenges at high speeds remains a future research topic. In addition, there is a requirement to bring electricity to the wheel and the need for wireless data transfer. As the major error source of MEMS gyros is eliminated, the method opens new applications for inertial navigation systems. In addition, there is a very large potential for wheel-based sensing in general, not restricted to Earth surface or navigation applications.
Published in IEEE Transactions on Vehicular Technology, June 2015.
The announcement expands the Leica Captivate Experience. With the addition of calibration-free GNSS and various upgrades to the Captivate field software and Infinity office software, users continue the immersive experience with self-learning GNSS and engaging, intuitive software interfaces.
