Precision Quantum Sensing

Established Quantum sensors measure time, gravity, acceleration, rotation, and magnetic and electric fields by reading out individual atoms, photons and spins, and they already hold every record that matters in those quantities. NIST’s explainer defines the field plainly: a quantum sensor uses quantum properties — quantised energy levels, spin, superposition, interference — to measure something in a way that would be impossible using classical physics alone. Its working modalities are atomic clocks, magnetometers, single-photon detectors, atom interferometers, electric-field sensors and squeezed light. The lineage is longer than most quantum technology: the first atomic clock ran in 1949, spin magnetometers date to the 1950s, superconducting SQUIDs to the 1960s, and magnetic resonance imaging — a quantum-spin measurement scaled to hospitals — to the 1970s.

Established Unlike quantum computing, quantum sensing is not a promissory field; it ships products and has done so for seventy years. Caesium-beam clocks have been commercial since the 1950s, satellite navigation is a constellation of flying atomic clocks, and chip-scale atomic clocks stable to roughly a part in 10¹¹ per day are catalogue items in telecoms and defence. The interesting economics of the field lie between the catalogue and the laboratory record, which in timekeeping are now separated by about eight orders of magnitude.

Frontier The live question of the 2020s is a transfer problem: how much of the laboratory record survives contact with a road, a ship, a hospital or an aircraft cabin. This brief tracks the digits, because in this field the digits are the argument: fractional frequency uncertainty for clocks (parts in 10¹⁸ and beyond), microgal-class accelerations for gravimeters (one microgal is about 10⁻⁹ of Earth’s surface gravity), femtotesla fields for magnetometers, and percent-of-distance-travelled error for navigation without satellites.

Established The best clocks in the world are now specified to nineteen decimal places. Caesium fountain primary standards, which still define the SI second, realise it at one to two parts in 10¹⁶. Optical clocks passed them nearly two decades ago and have kept going: in 2024 the JILA group at Boulder published a strontium lattice clock with a total systematic uncertainty of 8.1 × 10⁻¹⁹ (Aeppli et al., Physical Review Letters), and on 14 July 2025 NIST announced that its aluminium-ion clock had reached 5.5 × 10⁻¹⁹ (Marshall et al., Physical Review Letters), which NIST describes as 41 percent better than the previous record and 2.6 times more stable than any other ion clock. The stability gain is as consequential as the accuracy: extending each quantum probe from 150 milliseconds to a full second cut the averaging time needed to reach the clock’s floor from three weeks to about a day and a half. The instrument took twenty years of continuous development.

Established 2024 was the year the nuclear clock stopped being hypothetical. In April 2024 a PTB and TU Wien team (Tiedau et al., Physical Review Letters) achieved the first laser excitation of the thorium-229 nuclear isomer — the lowest-energy nuclear transition known, at about 8.4 electron-volts, reachable by a vacuum-ultraviolet laser near 148 nanometres — in thorium-doped calcium fluoride crystals. In September 2024, Jun Ye’s group at JILA, with TU Wien and LMU Munich collaborators, published in Nature the first direct frequency comparison between a nuclear transition and an atomic clock: a vacuum-ultraviolet frequency comb, coherently upconverted to its seventh harmonic, drove the thorium-229 transition in CaF₂ and measured its frequency ratio against a strontium-87 lattice clock, improving knowledge of the nuclear transition frequency by roughly a millionfold, to the kilohertz level, and resolving the nuclear quadrupole splittings. Nature’s own framing — that this marks the start of nuclear-based solid-state optical clocks — is for once not overclaiming.

Established Clocks are now geodetic instruments: general relativity makes every clock an altimeter. Gravitational time dilation shifts clock rates by about 1.1 parts in 10¹⁶ per metre of height near Earth’s surface, so a clock accurate to 10⁻¹⁸ resolves about a centimetre of elevation. In 2020 Takamoto, Katori and colleagues operated a pair of transportable strontium lattice clocks 450 metres apart at the Tokyo Skytree tower (Nature Photonics): an 18-digit frequency comparison that measured the gravitational redshift and confirmed general relativity’s prediction to (1.4 ± 9.1) × 10⁻⁵ — an order of magnitude better than any previous ground test and competitive with space experiments. An earlier transportable-clock campaign by PTB and INRIM at the Modane underground laboratory (Grotti et al., Nature Physics, 2018) had already demonstrated chronometric levelling in the field, at accuracy then corresponding to decimetres of height rather than centimetres.

Established Cold-atom gravity sensors have left the laboratory and found real infrastructure under real ground. In 2022 a University of Birmingham team (Stray et al., Nature) used an atom-interferometric gravity gradiometer — two clouds of about 10⁸ rubidium atoms, cooled to two to three microkelvin and separated vertically by one metre in an hourglass geometry so that shared lasers cancel vibration, tilt and laser noise — to detect a two-metre-square utility tunnel whose top sits about half a metre beneath a campus road. Ten measurements over about fifteen minutes sufficed to detect such a tunnel, with roughly half-metre spatial resolution. Project lead Kai Bongs was precise about the achievement and its limits: theirs was the first such sensor to work outdoors and still be sensitive enough to find tunnels, while running at roughly one-thirtieth the sensitivity of the best laboratory instrument. Commercial absolute quantum gravimeters (Exail, formerly Muquans) measure at the microgal level and have been deployed on volcanoes; marine campaigns with cold-atom gravimeters on hydrographic vessels have been reported by French groups since 2016.

Frontier Quantum navigation has produced its first at-sea and in-air trial numbers — all of them, so far, scored by the vendors themselves. In April 2025 the Australian firm Q-CTRL reported airborne trials of its Ironstone Opal magnetic-anomaly navigation system: on a roughly 500-kilometre fixed-wing flight, positioning uncertainty of about 0.03 percent of distance travelled, best runs near 0.01 percent (some tens of metres), a claimed advantage of at least 11 times over the paired inertial navigation system, about 6 times in ground-vehicle trials, and 99.97 percent in-flight uptime, using publicly available magnetic anomaly maps (arXiv:2504.08167) (vendor’s own figures). In August 2026 the same firm announced a maritime trial in the Coral Sea of gravity-map-matching navigation with a software-ruggedised quantum gravimeter installed in a passenger cabin, holding position error to about one nautical mile without satellite input and claiming more than tenfold improvement over a navigation-grade GNSS-backup system (vendor press release, not yet peer-reviewed). No independent standards body has yet adjudicated any of these claims.

Established Quantum magnetometry has produced in-vivo clinical results. Optically pumped magnetometers (OPMs) small enough to wear have been assembled into magnetoencephalography helmets. In a 2022 pilot study published in Radiology, a Brussels and Nottingham team fitted five children aged five to eleven with 32 OPM sensors mounted in customised caps: the wearable system identified interictal epileptiform discharges in all five, with peak amplitudes 2.3 to 4.6 times higher than a conventional 306-channel SQUID system — the OPMs sit about three centimetres closer to the brain — and signal-to-noise gains of 27 to 60 percent in four of the five children, at the price of generally noisier raw signals. Wearable OPM-MEG lets children move during scans, something cryogenic SQUID helmets cannot allow.

Established Squeezed light is in production use at kilometre scale. The LIGO gravitational-wave detectors have injected squeezed vacuum since 2019 and frequency-dependent squeezing since the O4 observing run began in 2023, buying several decibels of broadband quantum-noise reduction — the clearest demonstration that engineered quantum states of light improve a working scientific instrument rather than a demonstration bench.

Frontier Can entanglement improve a deployed sensor, not just a laboratory comparison? Spin squeezing and multi-ensemble entanglement have pushed laboratory optical-clock comparisons modestly past the standard quantum limit, and LIGO’s squeezer is an entanglement-adjacent device in production. But every gain demonstrated in matter-wave sensors so far is a few decibels under benign conditions; nobody has fielded an entangled gravimeter or clock outside a laboratory.

Frontier Does the nuclear clock beat the atomic clock, and by how much? The thorium-229 transition’s appeal is a solid-state host (no atom-by-atom trapping), a high line quality factor, and predicted sensitivity to variation of the fine-structure constant enhanced by factors in the thousands relative to atomic transitions — the enhancement itself is a model-dependent calculation, not a measurement. Against that stand crystal-field line shifts, temperature dependence, and radiation damage in the doped crystal, none yet evaluated at the 10⁻¹⁸ level. Trapped Th³⁺ ions are the cleaner but slower alternative path.

Frontier How far can clock comparisons push new-physics searches? The BACON collaboration’s clock network measured frequency ratios at 18-digit accuracy in 2021, and ratio measurements at the low 10⁻¹⁸ level now constrain present-day drift of fundamental constants and couplings of ultralight scalar dark matter. Every result so far is null; the science case is exclusion power, which grows directly with clock accuracy and comparison baselines.

Frontier When does chronometric geodesy beat conventional levelling? Clock networks promise geopotential differences over hundreds of kilometres without the error accumulation of spirit levelling, and could referee the decimetre-level disagreements between national height systems. The Skytree result shows centimetre capability over 450 metres; nobody has yet run a sustained clock-geodesy campaign against conventional methods over a 100-kilometre baseline.

Frontier Can frequencies be compared across oceans at 10⁻¹⁸? Optical fibre links compare clocks at the 10⁻²⁰ level across a continent, but the CCTF roadmap is blunt that no intercontinental link of remotely comparable quality exists; satellite techniques sit at 10⁻¹⁶ to 10⁻¹⁷. Free-space optical time transfer to satellites is the contested candidate fix.

Speculative Mid-band gravitational-wave detection with atom interferometers is a coherent proposal with no instrument yet capable of it. MAGIS-100 and the UK’s AION programme are explicitly pathfinders for a future kilometre-scale detector covering the 0.1–10 Hz band between LISA and LIGO; the pathfinders themselves are not expected to detect gravitational waves.

Speculative Large-scale atom interferometry doubles as a test of quantum mechanics itself — superpositions of atom clouds separated by metres for seconds probe decoherence and, in some proposals, gravity’s role in it. The proposals are serious; discriminating power at MAGIS scale remains to be shown.

Frontier The nuclear clock’s binding constraint is light: there is no continuous-wave laser at 148 nanometres. The JILA comparison ran on a vacuum-ultraviolet frequency comb generating microwatt-class harmonic power; clock operation wants narrow-linewidth, stable, preferably continuous excitation deep in the VUV, where mirrors, windows and nonlinear crystals all degrade. Laser development, not nuclear physics, sets the schedule.

Frontier Thorium-229 itself is a supply-chain oddity. Usable material is extracted from decades-old uranium-233 stocks held by weapons-legacy programmes, and worldwide research inventories are measured in micrograms to milligrams. Doped-crystal clocks ease the demand per device but do not create a supply.

Established Ultrastable reference cavities are the quiet bottleneck of every optical clock. Clock lasers are pre-stabilised on optical cavities whose thermal noise floors, vibration sensitivity and drift set achievable probe times; the state of the art uses cryogenic single-crystal silicon cavities that are neither transportable nor commercial. The record clocks’ second-long probe times lean directly on this infrastructure.

Established Reliability, not accuracy, is what currently disqualifies optical clocks from timekeeping duty. The CCTF redefinition roadmap records typical optical-standard uptimes ranging from a few percent to about 90 percent, against caesium fountains that run for months; and it lists the availability of commercial optical frequency standards as an unmet condition — none exist. A timescale needs a clock that is merely excellent every hour of the year, not perfect on its best day.

Established Atom interferometers pay for sensitivity with fragility on moving platforms. Free-falling atom clouds must be interrogated in a quiet inertial frame: platform vibration, rotation (which smears interference through the Coriolis effect), and the dead time between drops all bite hard at sea and in the air. The Birmingham instrument’s thirtyfold sensitivity concession relative to laboratory rigs is the honest price of working outdoors, and closing that ratio is the central engineering race in gravimetry.

Established OPM magnetometers only work near zero magnetic field. Wearable MEG therefore lives inside magnetically shielded rooms with active nulling coils; Earth’s field is about ten orders of magnitude larger than the femtotesla brain signals. Shielded rooms cost more than the sensors, and unshielded operation remains a research problem.

Frontier Size, weight and power separate every record from every application. A record clock occupies a laboratory; a transportable clock fills a trailer or a tower platform; navigation wants a shoebox on single-digit watts. Photonic integration of cooling and interrogation lasers exists at component level but not yet as a fielded, full-performance instrument.

Established Optical frequency combs underpin the entire field. The comb (Nobel Prize 2005, Hall and Hänsch) is what lets an optical frequency be counted, compared and — as in the thorium measurement, via seventh-harmonic upconversion — delivered to wavelengths where no laser natively performs. Every clock ratio and every redefinition criterion assumes combs as solved infrastructure.

Established Time and frequency transfer is a dependency with a published scorecard. The CCTF roadmap puts fibre-link performance at the 10⁻²⁰ fractional-frequency level with about 50 picoseconds time accuracy, against 10⁻¹⁶–10⁻¹⁷ and about a nanosecond for the satellite techniques that actually connect the world’s timing centres today. The gap between those two lines is the difference between having record clocks and being able to use them internationally.

Established Chronometric geodesy needs the geoid as much as the clock. Interpreting a clock-rate difference as a height requires local geopotential knowledge at matching accuracy; the roadmap scores this criterion as 70 to 90 percent fulfilled at the levels TAI calibration needs. Conversely, map-matching navigation depends wholly on the quality of prior gravity and magnetic anomaly maps — Q-CTRL’s trials ran on publicly available magnetic databases.

Frontier The component supply chain is thin. Ultra-narrow lasers, ULE and silicon cavities, VUV optics, low-noise microwave chains and vapour cells with controlled buffer gases are produced by a handful of vendors and national laboratories; several are single-source. Scaling from dozens of instruments to thousands is a manufacturing question nobody has yet been forced to answer.

Frontier The decisive experiment is to turn thorium-229 from a measured transition into a working clock: excite the nucleus on demand, lock a laser to it, publish a complete systematic error budget, and hold a year-long comparison against strontium and aluminium-ion clocks at or below one part in ten to the eighteenth. That single campaign would settle whether the solid-state host’s line shifts can be controlled, whether the predicted thousandfold sensitivity to fine-structure-constant variation can be cashed into the strongest drift limits ever set, and whether the nuclear clock is a metrological instrument or a beautiful spectroscopy result. Every ingredient now exists in at least prototype form, and the groups involved project integrated nuclear clocks this decade.

Established For the second’s redefinition, the experiment is already specified as a scoreboard. The CCTF’s mandatory criteria are numeric: at least three optical standards on the same transition, in different institutes, at or below 2 × 10⁻¹⁸; ratio measurements agreeing below 5 × 10⁻¹⁸, each measured at least twice by different institutes; continuity with caesium below 3 × 10⁻¹⁶; and monthly TAI calibrations near 2 × 10⁻¹⁶ from at least five optical standards for a year. As of the roadmap’s publication, fulfilment ranged from under 30 percent (validated ratios) to essentially complete (caesium continuity). Whether those percentages reach 100 by the 2030 General Conference is the cleanest yes/no test this field will produce.

Frontier Quantum navigation needs a blind, adversarially adjudicated trial. Every sea and flight result to date — including the Coral Sea gravimetric demonstration and the 0.03-percent-of-distance airborne magnetic navigation runs — was executed and scored by the vendor. A government-run trial with sealed truth data, published protocols and a conventional strategic-grade inertial system as control would either mint the industry or deflate it; defence customers are the only plausible sponsors.

Frontier MAGIS-100’s first science run will calibrate the entire mid-band programme. With its laser laboratory complete as of January 2026, atom sources due from Stanford in late 2026, installation through 2027 and commissioning from 2028, the 100-metre strontium interferometer will show what phase noise, atom flux and wavefront control a vertical-kilometre-class instrument can really deliver, and set its first ultralight dark-matter limits.

Frontier Chronometric geodesy needs its head-to-head: a 100-kilometre clock link levelled against conventional geodesy at the centimetre. The Skytree measurement was a single vertical baseline with the answer known in advance; the convincing demonstration is an inter-city campaign whose clock-derived height differences are published before the conventional survey is unsealed.

Frontier The field’s engineering programme is compression: taking instruments that fill rooms and rebuilding them as products that survive transport, temperature and neglect. The reference points are stark. The Birmingham gradiometer accepted a factor-of-thirty sensitivity penalty to work on a road; Q-CTRL’s maritime gravimeter is claimed to have run autonomously from a passenger cabin without temperature control or gyroscopic stabilisation (vendor); transportable lattice clocks have operated in a tower and an Alpine tunnel but at uncertainties one to two orders behind their laboratory parents.

Established Uptime engineering is measurable and mostly unsolved. A 99.97 percent uptime over one flight is a different claim from year-scale unattended operation, which is what timekeeping and navigation both demand; the CCTF’s few-percent-to-90-percent uptime range for optical standards is the honest baseline. Automated relocking, redundant lasers and health monitoring are unglamorous and decisive.

Frontier The nuclear clock’s engineering path runs through VUV photonics: raising harmonic comb power beyond microwatts, surfaces that survive 148-nanometre flux, and crystal growth that controls doping, strain and radiation damage in Th:CaF₂ well enough for a stated error budget.

Frontier Deployment mechanics matter as much as physics: magnetically shielded rooms sized for a moving child rather than a supine adult; gravimeter platforms that reject a ship’s heave spectrum; laser systems certified for aircraft power and vibration; and — for any future space clock — radiation-hard versions of all of it. Each is ordinary engineering; the combination is why the catalogue trails the record by orders of magnitude.

Established Precision sensing is the experimental arm of fundamental physics on this map. Clock redshift tests and matter-wave interferometry are currently the sharpest laboratory probes of general relativity’s interface with quantum mechanics, the territory surveyed in Quantum Gravity; and the squeezed-light and mid-band atom-interferometer programmes feed directly into Gravitational Wave Engineering, where LIGO’s operational squeezer is already quantum sensing at the kilometre scale.

Established Astronomy and deep-space operations are timing customers. Very-long-baseline interferometry of the kind behind Mega Telescopes is limited by station frequency standards, and navigation and ranging for Deep Space Communications improve directly with onboard clock quality — NASA’s Deep Space Atomic Clock flew a mercury-ion standard in 2019–2021 precisely to move navigation onboard.

Established The magnetometry lineage runs through superconductivity into neurotechnology. SQUID sensors — kin to the technologies in Superconducting Infrastructure — defined biomagnetism for forty years, and wearable OPM arrays now feed the non-invasive end of Brain-Computer Interfaces.

Frontier Navigation, geodesy and Earth observation are converging on the same instruments: the gravimeter that maps an aquifer is the gravimeter that navigates a submarine, which is why the field’s civil and defence programmes are so entangled.

Established Redefining the second is an institutional process with named committees and dated votes. The 27th CGPM (2022) endorsed the CCTF roadmap whose preferred scenario is a redefinition at the 29th CGPM in 2030, with the 2026 CGPM validating the plan; BIPM’s own FAQ describes the CCTF finalising a draft proposal, a choice among three options — a single optical transition, a weighted ensemble of transitions, or fixing a fundamental constant (currently impractical) — and warns that a 2030 failure defers the matter to 2034 or later.

Established A dozen national metrology institutes carry the record-setting load — NIST and JILA in the United States, PTB in Germany, NPL in Britain, SYRTE in France, RIKEN and NMIJ in Japan, NIM in China among them — and the redefinition criteria are deliberately written to require agreement across institutes, not supremacy by one.

Established Defence and national-programme money dominates the applied end. The UK’s National Quantum Strategy (2023) committed 2.5 billion pounds over ten years with sensing and timing as named pillars; MAGIS-100 is funded by the US Department of Energy, the Gordon and Betty Moore Foundation and UKRI, with the Imperial-led AION consortium (Imperial, Liverpool, Cambridge, Oxford; a 7.2-million-pound initial programme) as its UK counterpart; and quantum-navigation firms sell first to navies and air forces.

Frontier What is missing institutionally is an adjudicator for performance claims. Metrology has the BIPM; navigation has no equivalent body running blind trials, so vendor-scored demonstrations stand unrefereed — a gap that will matter the moment procurement decisions hang on them.

Established This field is dual-use at its core, and its biggest near-term driver is the prospect of war without satellite navigation. GNSS jamming and spoofing are now routine in conflict zones; quantum inertial and map-matching navigation is explicitly marketed as the answer, which means the same instruments that survey aquifers are procurement items for weapons platforms. The ethics are those of any positioning technology: it guides ambulances and missiles alike.

Speculative Claims that quantum gravimetry or magnetometry will make the oceans transparent to submarine detection are, on present numbers, unsupported — detection ranges scale poorly with source depth, and no published field result approaches operational relevance. The strategic-stability worry deserves monitoring rather than alarm; the burden of proof sits with the claimants.

Established The medical applications carry ordinary but real device-regulation questions. Wearable OPM-MEG’s clearest beneficiaries are children with drug-resistant epilepsy being evaluated for surgery — the Radiology pilot showed equivalent spike detection with stronger signals in five of five children — and the path from pilot to standard of care runs through regulatory clearance, reimbursement and shielded-room capital costs that will decide who gets access.

Established Subsurface surveying raises modest privacy questions with a physical ceiling. Gravity cartography can find voids, pipes and basements from public land, but as its own developers state, the resolution is far too low for imaging in any detailed sense; the realistic governance question is survey consent for critical infrastructure, not through-wall surveillance.

Established Precision timing is already civilisational load-bearing infrastructure, mostly invisibly. Power-grid synchrophasors, cellular networks, and financial-market timestamping all free-ride on GNSS time; national laboratories have repeatedly warned that a prolonged GNSS timing outage would degrade all three. Terrestrial optical timescales and fibre time distribution are the resilient alternative, and they are byproducts of the record-clock programme.

Frontier A planet instrumented with clocks and gravimeters would watch itself change in real time: aquifer drawdown, ice-mass loss, magma movement and tectonic strain all express themselves as gravity and geopotential changes. Satellite gravimetry sees these at hundreds-of-kilometres resolution today; ground quantum networks would add the local scale where decisions get made. That is a plausible two-decade build, not a certainty.

Established Redefining the second is a rare act of deliberate civilisational maintenance — replacing the foundation of every physical unit that references time, without any user noticing, on a supermajority vote of treaty states. Done in 2030, it will likely stand for a century; the caesium definition has already stood since 1967.

Horizons below track the redefinition process, the nuclear clock, and the migration of laboratory records into fielded instruments.

• 10 yr: Frontier The CGPM votes on an optical second in 2030 (slip to 2034 is the roadmap’s own fallback); first closed-loop thorium-229 clocks publish error budgets; MAGIS-100 and AION produce first dark-matter limits; certified quantum inertial-navigation units ride naval vessels as INS aids; wearable OPM-MEG clears regulators in multiple jurisdictions.
• 25 yr: Speculative Optical clocks dominate TAI and caesium fountains retire to secondary duty; intercontinental frequency comparison at 10⁻¹⁸ via optical satellite links; clock-geodesy networks monitor height systems and sea-level reference frames operationally; a kilometre-scale mid-band atom-interferometer detector is funded or definitively rejected on pathfinder results.
• 50 yr: Speculative Space-based optical or nuclear clock constellations provide autonomous deep-space navigation and geodesy; clock networks either detect ultralight dark matter or exclude it across broad mass ranges; nuclear-clock-grade drift limits on fundamental constants reach parts in 10²¹ per year.
• 100 / 250+ yr: Handwave Sensor networks at the quantum-gravity interface — instruments precise enough that spacetime’s own fluctuations, if any at accessible scales, become the noise floor — are a recurring theoretical hope with no current experimental purchase.

• Depends on nothing on this map, honestly: precision quantum sensing is an upstream capability. Its record instruments consume mature laser, vacuum and cryogenic engineering rather than results from other briefs here; the dependency arrows point outward.
• Requires (not on this map) a thorium-229 clock operated closed-loop with a published systematic error budget; a CGPM consensus in 2030 on which optical definition of the second to adopt; commercial optical clocks that run unattended for a year rather than an afternoon; a thorium-229 supply chain beyond microgram remnants of legacy uranium-233 stocks; intercontinental frequency transfer below 5 × 10⁻¹⁸ so distant clocks can actually be compared; and a procurement market for quantum inertial navigation whose performance numbers are set by independent blind trials rather than vendor scoring.
• Enables resilient positioning, navigation and timing without satellites; chronometric geodesy and real-time Earth monitoring; drift searches on fundamental constants and ultralight dark matter; the sensor floor for mid-band gravitational-wave astronomy; and the redefined SI second every other unit will silently inherit.
• Adjacent to Quantum Gravity, whose testable predictions largely wait on exactly these instruments, and Gravitational Wave Engineering, which already runs squeezed-light quantum sensing in production and would host any future atom-interferometric detector.

Handwave “The second has effectively already been redefined by optical clocks.” It has not. The definition remains caesium’s 9,192,631,770 Hz, the CCTF’s own scoreboard showed several mandatory criteria under 30 to 50 percent fulfilled, no consensus option had been chosen as of the roadmap, and the earliest possible vote is 2030 with 2034 as the acknowledged fallback.

Handwave “The nuclear clock already beats atomic clocks.” The 2024 breakthroughs measured the thorium transition to kilohertz precision — a millionfold improvement, and still roughly six orders of magnitude short of where strontium and aluminium-ion clocks operate. The nuclear clock’s case is its ceiling and its physics sensitivity, not its present performance; no thorium device yet keeps time at all in the operational sense.

Handwave “Quantum navigation will replace GPS.” The best vendor-reported field numbers — tens to hundreds of metres over a 500-kilometre flight, a nautical mile held at sea — are three to five orders of magnitude coarser than GNSS’s metres-to-centimetres. These systems bound inertial drift when satellites are denied; they do not compete with satellites that are present. The correct comparison class is the inertial navigation system, and even there the published head-to-heads are the vendor’s own.

Handwave “Gravity sensors will see everything underground, or track submarines from orbit.” The Birmingham team, in its moment of triumph, said the opposite: the resolution is too low for imaging. Gravity falls off steeply with distance and integrates over all mass; finding a known-class void half a metre down is a genuine feat that does not extrapolate to detailed subsurface pictures, and no published result supports operational submarine detection.

Frontier “Entanglement will deliver exponential sensor gains.” Quantum metrology’s hard ceiling is the Heisenberg limit — a gain scaling from √N to N in particle number, polynomial, not exponential — and decoherence claws back most of it in practice. Demonstrated squeezing gains in clocks and in LIGO are real and worth having: single-digit decibels, honestly counted.

Established “Clock tests keep almost finding cracks in relativity.” Every clock-based test to date is null: the Skytree redshift measurement agreed with general relativity to (1.4 ± 9.1) × 10⁻⁵, and constant-drift searches keep returning zero at tightening precision. That is the value — exclusion at eighteen digits — not a hint of new physics so far.

Established “Quantum sensing waits on quantum computing.” The two share suppliers and physics but not fates: sensors need no error correction and no million-qubit machine, which is exactly why they ship first. A world where quantum computing disappoints is a world where quantum sensing still redefines the second.

Speculative “A clock at 10⁻¹⁹ is precision for its own sake.” The rebuttal is written above in applications — geodesy at the centimetre, dark-matter exclusion, navigation without satellites — but the honest form of the claim survives in one respect: nobody has yet demonstrated a commercial or governmental service that requires the nineteenth digit, and the field’s civil economics beyond timekeeping remain projections.