ASTM wrote to the FAA complaining about RID GPS requirements in June 2021
The tip line has produced quite a scoop, two, in fact, a letter from the chair of the RID ASTM working group Gabriel Cox to the FAA telling them that the required GPS z-axis accuracy for the pilot’s/operator location, generally derived from their smartphone is difficult to achieve and one referencing that letter and giving it context.
I am adding these for you dear reader to make of what you will.
If you have an anonymous tip to add, about any RPAS related subject please feel free to add it at the link above. If it’s wordy, please break it up into parts of no more than 200 words.
Have a great weekend all!
The context first
The FAA is aware that the rushed rulemaking on Remote ID has caused actual technical issues.
Their leadership has received a memo from technical experts (see below) that the GPS vertical accuracy (designed for law enforcement) is technically impossible.
FAA never proposed GPS-only BRID and as such wasn’t aware that GPS have poor vertical accuracy. 15 feet (in the rule) is unachievable.
These experts have proposed 50 ft vertical accuracy. FAA has considered a “Technical amendment” without notice or comment. This process is normally reserved for administrative errors (i.e. “typos”).
This would be a serious abuse of the process.
The FAA leadership even considered getting access to highly accurate data that is prohibited for use (E911) for purposes other than true emergencies.
Also, the FAA has been pressuring Apple to change their software to accept both WiFi Beacon and NAN in iPhones.
I don’t understand the specifics, but right now apparently half the phones (either iPhone or Android) will not be able to see the BRID signal because these protocols are not interoperable – it’s like beta and VHS!
They don’t want this to come out for fear that they’ll be exposed for their sloppy technical research during rulemaking.
Gabriel Cox, Chair ASTM Remote ID Working Group
gabriel.c.cox@ June 28, 2021
To: Ben Walsh, Sam Sadeghi Federal Aviation Administration
Subject: The UAS Remote ID rule requirements relating to vertical accuracy of the location of Ground Control Station RIN 2130-AL31
The ASTM F3411 (Remote ID) working group is in the process providing the FAA with an updated Remote ID technical standard (F3411) and a Standard Practice Means of Compliance (MOC) document which we expect will be approved and used as a MOC for the Remote ID Unmanned Aircraft final rule.
Within the Remote Identification of Unmanned Aircraft Systems final rule, Section 89.310(h)(2) contains a requirement for the ground control station (GCS) location vertical accuracy to be within 15 feet (95% of the time).
The ASTM Remote ID working group under F38, has concluded this vertical accuracy requirement to be impractical using COTS components (L1 GPS w/SBAS) commonly used for location services on GCS of consumer UAS.
The ASTM workgroup analysis of components, other regulatory findings, GPS standards and test data lead us to a value of 50 feet (15 meters) as the lowest practical requirement for vertical accuracy. The ASTM analysis and references include the following:
1. The GPS Standard Positioning Service Performance Standard only guarantees 13m (average) and a 33m (worst site) of vertical accuracy (95%).
This standard has the following forward: This document defines the levels of performance the U.S. Government makes available to users of the Global Positioning System (GPS) Standard Positioning Service (SPS).
The GPS SPS Performance Standard underscores the U.S. commitment to cooperate with other global navigation satellite system and augmentation system providers to ensure compatibility and interoperability of GPS with emerging systems for peaceful, civilian worldwide use…
Office of the Department of Defense Chief Information Officer Attn: Assistant for GPS, Positioning and Navigation 6000 Defense Pentagon Washington D.C. 20301-6000 b. The following section represents quantifies the committed accuracy: 3.8.3 SPS Position/Velocity/Time Accuracy Standards The commitments for maintaining PDOP (Table 3.8-1), SPS SIS URE accuracy (Table 3.4.1), and SPS SIS URRE accuracy (Table 3.4-2) results in support for position/velocity/time accuracy standards as presented in Table 3.6-3.
When creating AC 922-001 2 Transport Canada performed a similar assessment and arrived at similar conclusions on vertical accuracy requirements. d. Based on the commitments above, the workgroup agrees, that committing to a more accurate figure than the commitment of maintainers of our GPS system would not be practical with the components available to manufacture consumer UAS.
Based on the average committed accuracy of 13m and worst of 33m, the workgroups is comfortable with committing that 15m (50 feet) vertical accuracy would be reasonably achievable in consumer UAS products.
It is well understood that GPS receivers may often experience greater accuracies not only for most datapoints being in the center of the Gaussian curve, but also because our GPS services often exceed their commitments. Nonetheless, the opinion of the workgroup is to stay within the committed values of the DOD rather than optimistically commit to our GPS system exceeding their commitments.
2. The manufacturer state capabilities for COTS components with the applied vertical accuracy conversions. COTS GNSS product datasheets often express only horizontal accuracy figures and they are typically expressed in terms of 50% Circular Area of Probability (CEP50).
Therefore, we must first
(a) determine what is commonly stated as the horizontal accuracy (CEP50), then
(b) convert this to CEP50 to R95 accuracy, then,
(c) convert the R95 horizontal accuracy to R95 vertical accuracy. a. Common representatives of COTS GNSS products used in modern UAS design are the Ublox M8 3 and the Quectel L70 4 both are L1 GPS modules with SBAS/WAAS support.
Both give horizontal accuracy values for 2.5m. Workgroup member SMEs generally agree that these datasheets values are generally optimistically tested in favorable conditions similar to manufacturer stated automobile gas mileage.
However, we will use this value for this exercise.
b. To convert the CEP50 values to R95 5 (95th percentile) values, we must perform a positive 2D Gaussian distribution (Rayleigh distribution) which is a ratio of 2.45/1.18 or 2.1. Multiplying (2.5m) by this ratio gives an R95 horizontal accuracy of 5.25m.
c. Then, to convert 5.25m (R95) horizontal accuracy to vertical accuracy, we use the GPS SPS Performance Standard 6 (table 3.8-2) that provides an SPS Availability Standard (95%) of 15m horizontal and 33m vertical, so taking the 33/15 ratio provides a horizontal to vertical conversion factor of 2.2. Thus, multiplying 5.25m (R95) horizontal accuracy value by 2.2 gives us about ~12m (~39+ feet) of converted database accuracy.
Given the commonly stated horizontal accuracy of about 2.5 (CEP50) in commodity L1 GPS products, which converts to 5.25m (R95), which converts to about 12m (R95) of vertical accuracy, and the field observations of the member SMEs, and given the GPS SPS performance commitments, the workgroup is comfortable that consumer UAS using these common components should be able to reliably build UAS that meet a vertical accuracy of 15m.
Smartphones as part of a GCS When using a smartphone as a GCS attachment, the industry cannot necessarily rely on the increased z-axis accuracies require by the new e911 FCC requirements.
The mandated requirements are to provide increase accuracy z-axis information in the location data used in the 911 calls and the use of this enhanced location information is highly restricted per the FCC rule 7.
However, since these handheld devices will generally have GPS for location services, we expect the performance of location services provided to third-party applications (such as GCS apps) to be capable of providing the recommended accuracy value of 15m (50 feet).
Regards,
Gabriel Cox Chair, ASTM Remote ID Working Group
1. https://www.gps.gov/technical/ps/2020-SPS-performance-standard.pdf
2. AC 911-001 (Draft), https://www.icao.int/safety/UA/UAID/Documents/DraftAC922-001.pdf
3. Ublox M8 Datasheet. https://www.u-blox.com/sites/default/files/NEO-M8-FW3_DataSheet_UBX-15031086.pdf
4. Quectel L70 Datasheet, https://www.quectel.com/wp-content/uploads/pdfupload/Quectel_L70_GPS_Specification_V2.3.pdf
5. Using Rayleigh Distribution table in “European Space Agency, Navipedia: Accuracy”,
6. Global Positioning System (GPS) Standard Positioning Service (SPS) Performance Standard – 5th Edition, April 2020, Table 3.8.2, SPS Position Availability Standards (https://www.gps.gov/technical/ps/2020-SPS-performance-standard.pdf)
7. 47 CFR 9.10(i)(4)(v) https://www.ecfr.gov/current/title-47/chapter-I/subchapter-A/part-9/subpart-C/section-9.10#p-9.10(i)(4)(v)
