BRENNEN

Industries / Construction and property / Engineering and architecture / Surveying

Mapping the ground

Mapping the ground is the work of fixing the position and height of points across a region or a whole country, so that every later survey, map and building can be tied to the same framework.

7

Milestones achieved, from 1617 to 2020.

1

To watch for: announced, not yet confirmed outside.

2

Open deliverables. Nobody has shipped them.

The road

AchievedWatch forOpen

Achieved

  1. 1617

    Triangles measure a country

    Measuring a long distance with chains is slow, and the small errors add up. Triangulation measures one short baseline with great care, then only the angles of a chain of triangles, and computes every other distance by trigonometry. Gemma Frisius described the method in 1533, in a booklet published with his Cosmographia, which the MacTutor history archive calls the first proposal to use triangulation to locate places accurately. Willebrord Snell, a professor at Leiden, used it in his book Eratosthenes Batavus (1617): from a measured baseline he built triangles across the Netherlands to find the distance between Alkmaar and Bergen-op-Zoom, about 130 km apart and nearly on one meridian, and from it the size of the Earth. He read his angles with a quadrant of about two feet (0.6 m) radius to the nearest minute of arc.

    mathshistory.st-andrews.ac.ukmathshistory.st-andrews.ac.uk

    Credit

    Gemma Frisius; Willebrord Snell, Leiden University

  2. 1784

    A baseline across Hounslow Heath

    In 1784 the Royal Society asked William Roy to connect the Royal Observatories of Greenwich and Paris by triangulation, to settle a dispute over their relative positions. Roy measured a five-mile (8 km) baseline on Hounslow Heath, a line that now crosses Heathrow airport, and the instrument maker Jesse Ramsden spent three years building a Great Theodolite, an instrument for measuring horizontal and vertical angles, three feet (0.9 m) across. The baseline later became the base of the Principal Triangulation of Great Britain, and Ordnance Survey counts 21 June 1791, when the Board of Ordnance bought a second Ramsden theodolite, as its founding date.

    ordnancesurvey.co.uk

    Credit

    William Roy; Jesse Ramsden; the Royal Society

  3. 1957

    Distance by microwave

    Trevor Wadley of South Africa's National Institute for Telecommunications Research designed the Tellurometer in 1957. A master unit sent microwaves to a remote unit at the far end of the line, which sent them back, and the instrument worked out the distance from the returning signal. The Science Museum Group says it measured long distances over rolling country to a few centimeters in about 20 km. The light-based Geodimeter NASM-2A weighed about 100 kg and could be used only in low light, such as at night; a Tellurometer MRA 101 of about 1962 weighed 7.3 kg and gave a reading in about a minute, by the account of the University of New South Wales instrument collection.

    collection.sciencemuseumgroup.org.ukvm.civeng.unsw.edu.auvm.civeng.unsw.edu.au

    Credit

    Trevor Wadley; South Africa's National Institute for Telecommunications Research

  4. 1968

    Angles and distance together

    A total station is a theodolite with a built-in electronic distance meter that records what it measures. Zeiss Oberkochen showed a prototype of its Reg Elta 14 at the German surveyors' Geodetic Day in Stuttgart in 1968. It measured distances up to 2,000 m with a standard deviation of 5 to 10 mm and punched angles and distances onto paper tape, so that field readings could pass to the drawing office without being copied by hand. It went into series production in 1970, and at the 1972 Munich Olympics it measured the length of throws in place of tapes. Sources differ on which instrument came first: a history published by the Alberta Land Surveyors' Association calls AGA's Geodimeter Model 700 of 1971 one of the first total stations, and Hewlett-Packard advertised its HP 3810A as the "New Total Station" in 1975.

    geodetski-vestnik.comalbertalandsurveyhistory.ca

    Credit

    Carl Zeiss Oberkochen; AGA Geodimeter; Hewlett-Packard

  5. 1983

    Satellites match the ground survey

    In January 1983 the Federal Geodetic Control Committee (FGCC), a group of US federal agencies, tested the Macrometer V-1000, a receiver that used signals from GPS satellites to find the position of one survey station relative to another without the two needing to see each other. The FGCC compared it with ground surveys tied to the US Transcontinental Traverse, a precise chain of measured lines and angles across the country. On lines 8 to 42 km long, the lengths agreed to a mean of 0.2 cm with a standard deviation of 1.7 cm, after allowing for a scale difference of 1 part in 455,000; on lines of 0.2 to 0.8 km they agreed to about 3 mm. The committee called it "a viable survey system" for national control.

    geodesy.noaa.gov

    Credit

    Macrometrics, Inc.; the FGCC Instrument Subcommittee (Larry Hothem and Charles Fronczak)

  6. 1990

    Trig pillars give way

    Between 1935 and 1962 Ordnance Survey retriangulated Great Britain from about 6,500 concrete triangulation pillars, built on hilltops so that surveyors at one pillar could sight lights on others. In 1990, 28 years after the retriangulation was finished, Ordnance Survey began a national GPS network to replace it.

    LessonA national network built for sight lines was overtaken within one working life once satellites removed the need to see from point to point.

    ordnancesurvey.co.uk

    Credit

    Ordnance Survey

  7. 2020

    Everest measured from the top

    The height of 8,848 m for Everest had been accepted since 1955. Nepal's Survey Department sent a team led by Khim Lal Gautam to the summit on 22 May 2019 with a GNSS receiver (GNSS covers GPS and the other satellite positioning systems), and New Zealand's government helped with the GNSS, leveling and gravity data. A Chinese team measured from the summit in May 2020 while a survey aircraft took gravity readings overhead. On 8 December 2020 the two countries announced a joint height of 8,848.86 m.

    nepalitimes.com

    Credit

    Survey Department of Nepal; Chinese surveyors; the New Zealand government

Watch for

  1. A crowd-built correction network

    RTK (real-time kinematic) positioning gives centimeter accuracy by streaming corrections from a base station at a known position. GEODNET, a network run on a blockchain with its own GEOD token, lets individuals and businesses buy certified base stations and add them to it; its creator Mike Horton wrote in February 2024 that it had more than 3,000 stations working. In May 2025 GEODNET and the module maker Quectel said the network had over 15,000 registered stations, every one using the same triple-band equipment, and that Quectel's modules with GEODNET corrections deliver centimeter-level accuracy for uses including surveying.

    Achieved whenan independent group publishes measured fix accuracy against surveyed marks for many GEODNET stations in more than one country, including how accurate the stations' own coordinates are.

    geodnet.cominsidegnss.com

    Claimed by

    GEODNET and Quectel, May 2025

Open

  1. 01

    Audit the volunteer base stations

    An independent check of the base stations that volunteers and small networks run for free or low-cost RTK corrections. For each station, log 24 hours of its public stream, compute its position separately with a government positioning service, and compare that with the position the station tells its users it is at. Then test fixes from a sample of the stations on published survey marks.

    The brief
    Credit · unclaimed
  2. 02

    Which survey marks still stand

    A model that predicts, for every survey mark in one county, whether it is likely still in place and usable, from its public record (when it was set, what it is set in, how to find it, and its recovery history) and recent aerial photos of the spot, followed by field checks of the predictions.

    The brief
    Credit · unclaimed

Open 01 · the brief

Audit the volunteer base stations

Credit · unclaimed

The first to publish is credited on this line, and the deliverable moves from Open to Achieved.

What to build

An independent check of the base stations that volunteers and small networks run for free or low-cost RTK corrections. For each station, log 24 hours of its public stream, compute its position separately with a government positioning service, and compare that with the position the station tells its users it is at. Then test fixes from a sample of the stations on published survey marks.

Why it is still open

A rover inherits any error in its base station's stated position, and these positions are set by the owners or the network. RTK2go, a free community caster, says it has more than 800 base stations online at any time and more than 11,000 registered; France's CentipedeRTK gives owners guides to compute their own base positions; GEODNET claims over 15,000 registered stations. This research found no independent audit of these stations' positions; the nearest study found, published in July 2025, compares RTK accuracy with several correction services, among them EUREF and EGNOS.

rtk2go.comdocs.centipede.frlibrary.kab.ac.ug

Built with what exists

RTKLIB's str2str to log streams and convbin to turn them into standard RINEX files, for stations that send raw observations; Natural Resources Canada's free CSRS-PPP service or the US National Geodetic Survey's OPUS for independent positions; a multi-band receiver such as a u-blox ZED-F9P for the field test; Python.

  1. First step

    Log 24 hours from five RTK2go streams near a government reference station, process the files through OPUS, and compare the results with the positions the streams broadcast.

  2. Done looks like

    For 100 stations in at least three countries, a public table of each station's stated position against its independently computed position, horizontal and vertical, in cm; and for 20 of them, 30 fixes each on a published mark within 20 km, with the share of fixes more than 5 cm off.

Open 02 · the brief

Which survey marks still stand

Credit · unclaimed

The first to publish is credited on this line, and the deliverable moves from Open to Achieved.

What to build

A model that predicts, for every survey mark in one county, whether it is likely still in place and usable, from its public record (when it was set, what it is set in, how to find it, and its recovery history) and recent aerial photos of the spot, followed by field checks of the predictions.

Why it is still open

The US National Geodetic Survey keeps records on more than 800,000 survey marks and relies on the public to report their condition through its Mark Recovery Form, mark by mark. Searches found no published model that predicts which marks have been lost, so surveyors and volunteers still search for each one blind.

geodesy.noaa.gov

Built with what exists

NGS datasheets, downloadable by county; free USDA NAIP aerial imagery; a vision-language model, an AI model that reads both text and images, to read each mark's written description and compare it with the imagery; a phone with a GNSS app for the field visits.

  1. First step

    Download one county's datasheets and count its marks by last reported condition and by the year of their last recovery.

  2. Done looks like

    Predictions for every mark in one county posted publicly before any field visit; at least 200 of those marks then searched for from public land or road verges away from traffic, each result reported to NGS; and a published table of predicted against found, with the share predicted correctly.