Third-party study finds that Swift’s Skylark Precise Positioning Service outperforms other corrections services by delivering better coverage and more consistent performance across the United States. Measured across four variables–coverage, accuracy, RTK fix rate, and convergence time–Swift performed better than every other service, making it the ideal GNSS corrections service for mass market deployments.
Location-based products are rapidly evolving to meet new market needs. From autonomous vehicles to augmented reality, and even robotic lawn mowers, technology that understands its location with ever-increasing precision is generating greater value each day. For instance, vehicle navigation systems–long able to tell you which lane you should be in for a freeway exit–can now tell you if you’re actually in the correct lane, providing better assistance to drivers and paving the way for autonomous vehicles.
Correcting errors in the GNSS signal to determine an object’s precise location isn’t a new concept either; the agricultural industry has been leveraging precise positioning to guide tractors for years. But as the demand for precise positioning increases across industries, this technology, too, is rapidly evolving. RTK positioning–which determines an object’s exact location in small, defined areas, where a local RTK base station can serve as a source of truth for the relative location of other devices–has given rise to cloud-based precise positioning, which can determine the precise location of millions of devices traveling at all speeds and in all directions across continents and around the world.
Several corrections services have emerged to enable the new wave of mass market applications. But how do these services perform in real world environments? And how do they compare to one another?
Hyfix Spatial Intelligence conducted a study to determine which service provides the best performance in the United States. They compared three mass market corrections networks (Swift Navigation’s Skylark Precise Positioning Service, Point One Polaris, and Verizon Hyper Precise) with a traditional survey-grade service used in agriculture (Hexagon SmartNet). Notable among these services is that only Skylark is a PPP-RTK service, while Polaris, Hyper Precise, and SmartNet are RTK services. RTK services are dependent on proximity to local base stations, but PPP-RTK services can deliver uniform accuracy country-wide.
Hyfix used National Geodetic Survey station logs and the open source positioning engine RTKLIB for a fair comparison across all corrections providers. They analyzed each service across four variables:
To evaluate each service’s coverage, Hyfix selected 35 locations across the country, in a mix of both urban and rural areas. Unsurprisingly, Hexagon SmartNet trailed its mass market competitors in this dimension. Because SmartNet was designed to cater to traditional agricultural use cases, base stations are strategically placed in particular areas in the country, leaving many regions without coverage. Among the other services, only Swift’s Skylark successfully generated corrections in every single location. Because Point One Polaris and Verizon Hyper Precise are dependent on proximity to local base stations, the only way to improve their coverage is to stand up additional base stations, but Skylark works everywhere.
Winner: Swift
Hyfix measured horizontal and vertical accuracy of each corrections service. In this case, we would expect the traditional survey-grade service to set the standard for performance. Indeed, Hexagon SmartNet performed exceptionally well, with horizontal and vertical accuracy of 5 cm. But two of the mass market services performed comparably, with both Swift and Verizon producing similar results. Swift eked out Verizon in the accuracy category with a smaller standard deviation, meaning that it delivers high accuracy results with less variance.
Tie: Hexagon and Swift
When each corrections service provides corrections to RTKLIB, the positioning engine calculates the final results. RTK fix rate measures the number of positions in the session that resolved to an integer fix, the highest level of accuracy. For pure RTK services, the quality of the position data degrades the farther a device is from the local base station, resulting in a lower RTK fix rate and with that, lower accuracy.
Only Swift had a near-perfect RTK fix rate, meaning that the accuracy delivered by Swift is consistent and reliable across the country. Other services had significantly lower RTK fix rates, making them less dependable services for mass market deployments.
Winner: Swift
In ideal conditions, a location-enabled device will converge on an RTK fix very quickly. The time it takes to converge is dependent on both the quality of the corrections and also the quality of the positioning engine. With coverage gaps or a degraded GNSS signal, the positioning engine may never converge to an RTK fix. When that happens, it will continue to try to resolve its RTK position until it times out, draining battery life in the process.
RTKLIB enables convergence to an RTK fix in just a couple seconds with good corrections data. In the Hyfix study, Swift dramatically outperformed the other corrections services in convergence time, largely due to the consistent quality and coverage of the corrections, enabling RTKLIB to resolve to an RTK fix almost immediately 99% of the time. Other services often resulted in RTKLIB being unable to resolve the RTK position, which impacted their average convergence time. In real world use cases, this means that Swift’s corrections will reliably converge as fast as the positioning engine allows, wherever the device is, while other services may only converge quickly in ideal conditions.
Winner: Swift
None of these variables can be considered in a vacuum. Together, coverage, accuracy, RTK fix rate, and convergence time determine the overall viability of a corrections service for mass market deployments. Coverage should be considered first, with the simple question of “does the service successfully generate corrections everywhere my customers are?” Only Swift was able to generate corrections in every location sampled in the Hyfix study.
From there, accuracy, RTK fix rate, and convergence must be evaluated together. RTK services will provide a greater level of accuracy and fast convergence times when your customers are close to a base station, but you cannot always depend on ideal conditions for a mass market deployment. Although a PPP-RTK service may deliver slightly lower accuracy in some locations, averaged across a large number of devices in real world scenarios, the accuracy is equivalent to RTK, with consistent convergence times wherever your customers are. As demonstrated in Hyfix’s data, Skylark delivered as good or better accuracy than RTK services when averaged across the 35 locations and fast convergence times everywhere–exactly what is needed for mass market deployments.
If you’re deploying mass market location-enabled devices across the country, you need confidence that those devices will perform as expected wherever they are. Unlike high-end agricultural machinery that is sold to sophisticated buyers in known, fixed locations, mass market products must be able to weather the uncertainty that comes when you don’t have total control over where your devices will be operating.
Only Swift Navigations’s Skylark Precise Positioning Service delivers the reliability and consistent performance across the United States required to support mass market applications.
Bonus: for use cases where you require 2-4 cm accuracy and your devices operate in constrained locations, Swift also offers an RTK service.
Want to learn more about Skylark? Contact Swift to get started.