PSM SEPTEMBER 2013
In our September issue PSM reports from the PPP symposium held in Toronto, Canada by the International GNSS Service.
Global and Real-Time
PSM began our in-depth coverage of Precise Point Positioning in 2012, continuing through 2013. PPP’s technology has been around in various formats for nearly two decades in academic circles, and recently it has entered a bit of a renaissance period. Our prediction appears to be on track that PPP and PPP-RTK will become a viable successor (or at least an essential augmentation) to RTK and RTN.
In the September 2013 issue of PSM, Grant Hausler, a Ph.D. candidate from down-under, answered the question—“Is PPP just the flavor of the month?”—with insights from the recent PPP symposium held in Toronto, Canada by the International GNSS Service (IGS). The IGS is the international cooperative developing and maintaining the authoritative world geodetic reference framework, working with dozens of scientific bodies, academic institutions, and government entities.
PPP is an alternate method for GNSS positioning that relies heavily on high-quality clock and orbit “products.” Some implementations enable users to produce geodetic positions without bases, without networks, without cellular service. You can use post-processed PPP today, anywhere in the world; you just need a receiver that can log Receiver Independent Exchange Format (Rinex) and upload it to sites like those hosted by NASA/JPL and Natural Resources Canada (NRCAN) to get automated results—sort of like a “global OPUS.”
While the developments by the IGS and academia are noteworthy, those happening in the commercial and open-source communities may affect surveyors more rapidly and directly.
Commercial developers, especially for agriculture and marine markets, have been using PPP and variants for over a decade. John Deere has Starfire for precision agriculture. Veripos and Fugro have been delivering their real-time PPP solutions via satellite for marine customers. Surveying users have seen a taste of “PPP-RTK by-products” in some rovers, like the xFill on the Trimble R10 rover and a Veripos solution announced for the Altus APS-3L. GEO++ has long used elements of PPP-RTK as the foundation of its network solutions. The heartbreak of PPP for surveying continues to be the long-convergence times needed to achieve viable precisions.
This is about to change.
The PPP Party Is Just Getting Started
Many of the existing real-time PPP services are targeted at precision agriculture and marine markets. More recently, some developers are aiming to serve, with PPP-RTK, many of the same end-user segments as are currently served by RTK and RTN infrastructure solutions. A growing number of precision agriculture users want RTK-quality GNSS, but to do that with current PPP-RTK methods there would need to be dense clusters of CORS (though not as dense as for an RTN). Real-time communications can be a serious challenge in rural areas as well.
In a move that caught many in the industry by surprise, Trimble not only jumped into the agriculture-centric “DGPS-quality” PPP services market but have also added an array of higher precision versions of real-time PPP. We asked Patricia Boothe, general manager of Positioning Services at Trimble about the future of their CenterPoint RTX services and to see what the PPP prospects are for surveyors.
“We look at CenterPoint RTX as a backbone of solutions and applications for positioning,” said Boothe. “Precision agriculture has used a number of correction services: DGPS beacons, SBAS (satellite-based augmentations systems, a lot like WAAS and some commercial services), RTK bases, and VRS.”
Boothe explained that Trimble has built their own worldwide tracking network to develop clock and orbit products for real-time PPP. These are delivered by either satellite or cellular and include a service with standard convergence times of about 30 minutes that can yield 4cm precisions. But, by adding data from a sparse network of CORS at around 100km spacing (as has been built by Trimble in the Midwest), a convergence time of one minute can yield 4cm precisions. The latter starts to sound like something surveyors could use (for some of their needs).
But, if an RTN (albeit with a tighter CORS spacing of <70km) can provide this already, what would be the advantage of a PPP-RTK solution? “The solutions can also be broadcast from a satellite; you do not have to deal with cell coverage issues,” said Boothe. “We see the ‘fast’ RTX as being a potential complement to areas served by [VRS] networks.”
He also noted that at the October 2013 Intergeo (international conference and exhibition for geomatics), Trimble announced full implementation of RTX for surveying rovers, like the R10 and new V10 Imaging Rover, the Pro XRT, and that many have been trying out the RTX online post-processed-positioning service. PPP has indeed hit the mainstream. Goodbye “cell hell”?
How Bright Is PPP’s Future?
By Sunil Bisnath
What is holding most current methods for PPP back from wider acceptance for other uses is the convergence period that solutions must go through. These periods, on the order of up to tens of minutes, are required for the PPP filter to converge from a pseudorange-based solution to a carrier-phase one. If too many satellite signals are blocked during operation, the position solution must be re-initialized from a few-meter-level, pseudorange-based solution. This practical limitation results in PPP being commercially viable in open sky areas where baseline GNSS or network RTK are not as physically or economically feasible.
A great deal of research focus is currently being placed on PPP by industry, academia, and governments (for example, see yorku.ca/pppworkshop2013). We have already seen improvements in error modeling, resolution of carrier-phase ambiguities, and the addition of GLONASS to GPS PPP. Significant improvements include: initial convergence period reduction from 30-45 minutes to 15-30 minutes (depending on data quality and the definition of converged) and re-initialization after few-minute measurement outages as opposed to complete re-initialization. Some implementations of real-time PPP, with additional CORS augmentation, seek to reduce convergence times even further.
Over the next few years, with additional constellations and signals, simulation studies and limited testing with real measurements point to significant further reduction in initial convergence period and small improvements in steady state positioning accuracy.
Will PPP replace network RTK? Well, additional measurements that will dramatically improve PPP will also make improvements to network RTK and baseline GNSS, albeit by smaller amounts as there are few remaining errors to model. So, while PPP will continue to approach RTK-like performance, it will be difficult to completely emulate RTK performance.
While it’s always difficult to try to predict the long-term future, beyond this decade it may be possible for multi-constellation PPP, combined with the next generation of low-cost, high-performance, multi-frequency GNSS receiver chips and antennas, low-cost, high-performance, MEMs-based inertial sensors, and advanced error modeling and mitigation techniques to provide low-cost satellite positioning and navigation.
Such a solution could be seen as the evolution from the current mass commercial market of low-cost, pseudo-range-based point positioning GPS receivers to low-cost, pseudorange and carrier-phase PPP GNSS chips integrated with other sensors in the spectrum of mobile devices.
Dr. Sunil Bisnath is an associate professor in the Department of Earth and Space Science and Engineering at York University in Toronto, Canada, was host of the June 2013 PPP symposium, and is a contributor to this magazine.