Quantum sensing (magnetometers, gravimeters, inertial navigation)

Scope and definitions

Quantum sensing here covers three instrument families:

1. Quantum magnetometers — optically pumped atomic magnetometers (OPMs), including spin-exchange relaxation-free (SERF) instruments, and diamond nitrogen-vacancy (NV) center magnetometers.
2. Quantum gravimeters/gravity gradiometers — cold-atom interferometers that measure gravitational acceleration or its spatial gradient.
3. Quantum inertial navigation — atom-interferometric accelerometers and gyroscopes, and hybrid classical-quantum inertial measurement units (IMUs).

A peer-reviewed review defines the field as sensors that use quantum coherence, entanglement, or discrete energy levels to transduce a physical quantity [S1]. Optical magnetometry is covered by a separate Tier-1 review [S2].

The load-bearing engineering distinction is: a granted patent records claims; it is not proof of performance, yield, cost, or commercial viability. Likewise, a journal paper usually reports laboratory or field-trial results under specific conditions, not an industrial product specification.

Technical foundations

2.1 Serf and optically pumped magnetometers

Optically pumped magnetometers polarize alkali-metal vapor with a laser; the spin precession frequency in an external magnetic field is read out optically [S2]. At high alkali density, spin-exchange collisions usually broaden the magnetic resonance. Allred et al. showed that operating in the SERF regime suppresses spin-exchange relaxation, allowing narrow linewidths [S3]. Kominis et al. then reported a multichannel SERF magnetometer with sensitivity below 1 fT/√Hz [S4]. The exact quoted sensitivity was 0.54 fT/√Hz for a single channel in a shielded environment [S4]. This number is cross-verified in the review literature as representative of SERF-class performance [S1][S5], but the original measured value is from one PRL/Nature paper lineage: Romalis group Princeton and co-authors [S3][S4].

SERF magnetometers require near-zero magnetic field at the vapor cell, so fielded instruments typically operate inside magnetic shields or use closed-loop field compensation. That is a major contradiction with “drop-in navigation sensor” marketing claims: sub-femtotesla sensitivity is achieved in magnetically quiet, often shielded, environments [S3][S4][S5].

2.2 Diamond NV-center magnetometers

NV centers in diamond are point defects that can be initialized and read out optically; their electron spin is sensitive to magnetic fields at room temperature [S6][S7]. The value proposition is spatial resolution: an NV center can sense nanoscale magnetic fields because the spin is localized in a diamond lattice [S6][S7]. A 2020 Reviews of Modern Physics paper provides the sensitivity-optimization framework for NV-diamond magnetometry [S5].

The trade-off is sensitivity versus volume. A single NV or small ensemble has high spatial resolution but much lower volume-averaged sensitivity than SERF OPMs. The review explicitly frames sensitivity optimization as a function of collection efficiency, coherence time, and NV density [S5]. This is a contradiction to “NV mags are always the best” claims: the same Tier-1 source shows that sensitivity scaling depends on exactly which resource is optimized [S5].

2.3 Atom-interferometric gravimeters and gravity gradiometers

Atom interferometry uses laser pulses to split, redirect, and recombine the wave packets of laser-cooled atoms. The interference phase shift depends on inertial acceleration [S8]. Reviews of cold-atom inertial sensors state that these devices provide absolute acceleration/rotation references with low long-term drift, at the cost of size, vibration isolation, and finite dynamic range [S9].

A 2018 Scientific Reports paper reported a transportable absolute quantum gravimeter achieving gravity measurements below \(10^{-9}\,g\) [S11]. A 2022 Nature paper demonstrated a gravity gradiometer used for “gravity cartography” — detecting an underground tunnel by mapping gravity-gradient signatures [S10]. These are independent institutional results: French metrology/geophysics groups for the transportable gravimeter [S11] and UK Birmingham-led collaboration for gravity cartography [S10].

2.4 Inertial navigation

Classical inertial navigation uses ring-laser gyros, fiber-optic gyros, and MEMS sensors to measure angular rate and acceleration. Atom-interferometric inertial sensors aim to remove drift-scale errors by referencing acceleration and rotation to an atomic transition [S9]. Geiger et al. describe high-accuracy cold-atom inertial measurements but note that compactness and dynamic range remain engineering bottlenecks [S9].

Worked quantitative calculation

3.1 Projection-noise-limited SERF magnetometer sensitivity

The spin-projection-noise-limited sensitivity of an optically pumped magnetometer can be written as:

\[

\delta B = \frac{1}{\gamma_e \sqrt{N T_2 \tau}}

\]

where:

• \(\gamma_e = g_e \mu_B / \hbar \approx 1.76 \times 10^{11}\ \text{s}^{-1}\text{T}^{-1}\) is the electron gyromagnetic ratio.
• \(N\) is the number of polarized alkali spins in the sensed volume.
• \(T_2\) is the transverse spin coherence time.
• \(\tau\) is the total measurement time; for \(\tau = 1\ \text{s}\), the result is a per-root-hertz sensitivity estimate.

Inputs used here:

• \(N = 10^{15}\) spins — an assumed, order-of-magnitude value for a dense alkali vapor cell of roughly 0.1–1 cm³; this is not a directly quoted cell volume from [S4], but an inferred design parameter from the SERF literature [S4][S5].
• \(T_2 = 100\ \text{ms}\) — same order as coherence times in the SERF regime; assumed for this calculation, not a certified product spec [S3][S4].
• \(\tau = 1\ \text{s}\).

Then:

\[

\sqrt{N T_2 \tau} = \sqrt{10^{15} \times 0.1 \times 1} = \sqrt{10^{14}} = 10^7

\]

\[

\delta B = \frac{1}{(1.76 \times 10^{11})(10^7)}

= 5.68 \times 10^{-19}\ \text{T}

= 0.568\ \text{fT}/\sqrt{\text{Hz}}

\]

This calculated projection-noise estimate is 0.57 fT/√Hz, which is close to the measured 0.54 fT/√Hz reported by Kominis et al. [S4]. The match should be interpreted cautiously: the inputs \(N\) and \(T_2\) are chosen to fit the SERF operating regime, not independently metered from the original paper in this note.

Sensitivity to inputs:

• If \(T_2 = 10\ \text{ms}\), then \(\delta B \approx 1.8\ \text{fT}/\sqrt{\text{Hz}}\).
• If \(N = 10^{14}\), then \(\delta B \approx 5.7\ \text{fT}/\sqrt{\text{Hz}}\).
• If readout noise or spin-destruction collisional noise dominates, the sensitivity degrades further [S1][S5].

3.2 Atom-interferometer accelerometer scale factor

For a Raman-pulse atom interferometer, the acceleration-induced phase shift is approximately:

\[

\Delta\phi = k_{\text{eff}} a T^2

\]

where:

• \(k_{\text{eff}}\) is the effective two-photon wave vector.
• \(a\) is the acceleration.
• \(T\) is the pulse-separation/interrogation time.

For rubidium-87 near 780 nm, a Raman transition can use an effective wave vector:

\[

k_{\text{eff}} \approx \frac{4\pi}{780\ \text{nm}}

\approx 1.61 \times 10^{7}\ \text{m}^{-1}

\]

For \(T = 100\ \text{ms}\):

\[

k_{\text{eff}} T^2

= (1.61 \times 10^7)(10^{-2})^2

= 1.61 \times 10^3\ \text{s}^2/\text{m}

\]

If the interferometer phase resolution is assumed to be \(\delta\phi = 1\ \text{mrad/√Hz}\), then:

\[

\delta a = \frac{\delta\phi}{k_{\text{eff}} T^2}

= \frac{10^{-3}}{1.61 \times 10^3}

\approx 6.2 \times 10^{-8}\ \text{m}/\text{s}^2/\sqrt{\text{Hz}}

\]

Since \(1\ \text{Gal} = 10^{-2}\ \text{m}/\text{s}^2\), this is:

\[

\delta a \approx 6.2\ \mu\text{Gal}/\sqrt{\text{Hz}}

\]

This is a derived order-of-magnitude estimate, not a measured product specification. The source-basis is the atom-interferometer formalism [S8][S9]. Transportable gravimeter results in [S11] show the same technology class can reach geophysically useful sensitivities, but the exact phase noise and \(T\) are deployment-specific.

Parameter table

Patents and IP

A granted patent relevant to NV-diamond magnetometry is:

• U.S. Patent 8,547,090 B2, granted 2013-10-01, available at Google Patents: [S12]. The record discloses and claims magnetic sensing using nitrogen-vacancy centers in diamond. The patent is cited here as evidence of recorded IP claims, not as proof of sensitivity, yield, or deployed product performance.

The IP moat question must be separated into three distinct sub-questions:

1. Does a granted patent show technical feasibility? No. A patent shows that an examiner accepted the claims as novel, useful, and non-obvious over prior art [S12].
2. Does a granted patent show commercial viability? No. Commercial viability requires cost, reliability, manufacturing yield, and buyer adoption; none of those are proven by a patent document [S12].
3. Does a granted patent show freedom to operate? No. Competing patents may still block practice.

Many quantum-sensing startups are private, so their patent portfolios are visible but their financial performance is not. That makes moat analysis intrinsically needs primary source validation on the revenue side.

Company disclosures and cross-reference

6.1 Honeywell

Honeywell markets inertial navigation hardware. The HGuide n500 inertial navigation system is a company product page with self-reported specifications [S14]. Honeywell files as a publicly listed company; its 10-K filings are accessible through SEC EDGAR [S13]. In those filings, inertial navigation and aerospace guidance products sit inside the Aerospace segment. No separate quantum-sensing revenue line is apparent in the 10-K filing listing [S13]. The company product page is Tier 2 self-reported marketing; the 10-K is Tier 1 regulatory disclosure. Observed disclosure gap: the HGuide page [S14] presents navigation performance claims, while the SEC filing search [S13] does not supply independent, third-party validation of quantum-grade navigation performance. This is not necessarily a contradiction, but it is a lineage mismatch: marketing claims should not inherit Tier-1 status from a 10-K that does not separately quantify them.

6.2 Defense primes

Lockheed Martin [S15], Northrop Grumman [S16], RTX [S17], L3Harris [S18], General Dynamics [S28], and BAE Systems plc [S27] are all publicly listed defense and aerospace firms with SEC filings. Several market inertial sensors, navigation systems, guidance units, or gravitational/geophysical instrumentation. Cross-referencing individual marketing pages against 10-K risk factors is a second-pass target; the key disclosure check is whether any of these companies reports a separate quantum-sensor revenue segment. None of the EDGAR 10-K listings reviewed here indicates a standalone quantum-sensor P&L segment [S13][S15][S16][S17][S18][S27][S28].

6.3 Component and clock makers

Microchip Technology [S20] and Frequency Electronics [S21] supply atomic clocks and frequency references used in navigation and timing. Teledyne [S19] supplies precision instrumentation. KVH Industries [S29] supplies fiber-optic gyro-based navigation. These are adjacent to, but not necessarily inside, quantum sensing. Listing them as quantum-sensing tickers without a confirmed quantum magnetometer/gravimeter/atom-interferometer product would overstate exposure. Their role here is value-chain adjacency, not pure-play quantum sensing.

6.4 Marketing versus filing discrepancy summary

• Claim: “Quantum sensors are ready for navigation.”

Filing evidence: Defense primes disclose navigation products broadly but do not disclose a separate quantum-sensor segment in the EDGAR 10-K listings reviewed [S13][S15][S16][S17][S18].

Assessment: needs primary source validation before assigning a revenue line.

• Claim: “SERF magnetometers deliver sub-femtotesla sensitivity.”

Tier-1 evidence: Yes, 0.54 fT/√Hz measured in a shielded laboratory configuration [S4].

Contradiction: The same Tier-1 literature shows SERF operation requires near-zero field and/or shielding, so the number is not an unshielded fielded-vehicle specification [S3][S4][S5].

• Claim: “NV diamond magnetometers can replace SERF sensors.”

Tier-1 evidence: NV centers enable nanoscale spatial resolution at room temperature [S6][S7], but sensitivity optimization is resource-dependent and does not automatically exceed SERF volume sensitivity [S5].

Assessment: The products serve different measurement niches; direct replacement is not supported by Tier-1 sources.

• Claim: “Atom interferometers can navigate without GPS.”

Tier-1 evidence: Cold-atom sensors offer low-drift absolute inertial references [S9], but dynamic range, size, and vibration issues remain [S9].

Assessment: Navigation-grade field use is still an engineering and procurement question, not a settled product category.

Unit economics / business model

Who monetizes:

• Defense/aerospace primes monetize through system integration contracts, navigation units, and long-term service/upgrade programs. SEC filings show U.S. government agencies are material customers [S13][S15][S16][S17][S18]. Margins on these programs are not disclosed at the quantum-sensor level.
• Private sensor startups monetize through hardware sales, development kits, pilot deployments, and government research contracts. Tier-2 company pages illustrate the model: SBQuantum describes diamond magnetometer hardware and geophysical sensing [S23]; Q-CTRL describes quantum control software and sensing solutions [S22]. These are self-reported and do not provide audited margin data.
• Software/control layer — Q-CTRL sells software-defined quantum control and sensing analytics [S22]. This is a software/hardware hybrid model; no public margin data were found in this pass.
• Component suppliers monetize through atomic clocks, frequency references, and precision timing [S20][S21]. Margins are disclosed only at aggregate segment level, not per quantum-sensor part.

Margin statement: No reliable, source-backed quantum-sensor gross-margin figure was identified in this research pass. Any specific margin number would be needs primary source validation.

Value capture: The value likely accrues to integrators who solve the hard fielded-environment problems — vibration, shielding, dynamic range, calibration — rather than to bare sensor heads. This is inferential from the Tier-1 sources showing that sensitivity degrades outside controlled conditions [S3][S4][S5][S9].

Disconfirming Evidence

The thesis that quantum magnetometers, gravimeters, and inertial navigation are commercially decisive would be weakened or falsified if:

1. SERF magnetometers cannot maintain sensitivity outside magnetically shielded laboratory environments; the Tier-1 measured 0.54 fT/√Hz is itself from a shielded configuration, and fielded sensitivity may be substantially worse [S3][S4][S5].
2. Atom-interferometric inertial sensors remain rack-sized, vibration-sensitive, and limited in dynamic range, preventing adoption in aircraft, submarines, or tactical systems [S9].
3. No public company filing discloses a material quantum-sensing revenue line or unit volume; if 10-K/10-Q filings continue to show quantum sensing only as R&D or as a small sub-component of navigation systems, the investment case is not a product-market case [S13][S15][S16][S17][S18].
4. Granted patents expire or fail to produce license revenue, and private startups continue to rely on government grants rather than repeat commercial buyers [S12][S22][S23].
5. Classical alternatives — MEMS IMUs, ring-laser gyros, superconducting gravimeters, or improved GPS-denied navigation techniques — remain cheaper, smaller, or sufficiently accurate for the same buyers [S9].