Quantum Parcels: Diverse Quantum Carriers

• Quantum Parcels are heterogeneous quantum objects defined as finite-precision state sets, localized wave packets, or transportable entanglement resources.
• They span multiple domains including interval quantum mechanics, discrete-time quantum walks, superfluid turbulence, and quantum communication.
• Quantum Parcels also underpin quantum optimization models, aiding in efficient package delivery and advanced quantum networking strategies.

“Quantum parcels” is not a single standardized term in the arXiv literature. It denotes, in different contexts, finite-precision sets of quantum states, localized wave packets engineered for storage and transport, passive tracers in quantum fluids, reusable entanglement carriers, containerized fault-tolerant memories moved as cargo, EPR packets buffered in quantum memories, and parcel-delivery optimization problems attacked with quantum algorithms. This plurality is explicit across Interval Quantum Mechanics, discrete-time quantum walks, Husimi-phase-space dynamics, superfluid turbulence, quantum networking, and quantum logistics (Edalat, 19 May 2026, Vieira et al., 2021, Zhdanov et al., 2021, Tang et al., 1 Apr 2025, Devitt et al., 2014, Valentini et al., 15 Apr 2026, Osaba et al., 2024, Moosavi et al., 7 Aug 2025).

1. Terminological scope and major usages

The literature uses the phrase in at least six technically distinct ways. In some papers the term is literal and ontological, while in others it is operational or metaphorical. This suggests that “parcel” usually marks a localized, transportable, finitely specified, or resource-bearing quantum object, but the underlying mathematics and physical interpretation vary substantially.

Domain	Meaning of “quantum parcel”	Representative source
Interval Quantum Mechanics	weak$^*$-open convex set of density matrices defined by finitely many expectation intervals	(Edalat, 19 May 2026)
Superfluid turbulence	passive tracer advected by the instantaneous superfluid velocity field	(Tang et al., 1 Apr 2025)
Husimi dynamics	elementary parcel of Husimi fluid centered at $(p_c(t),x_c(t))$	(Zhdanov et al., 2021)
Discrete-time quantum walks	robust, practically dispersionless Gaussian qubit packet	(Vieira et al., 2021)
Quantum communication	fixed entangled state used repeatedly to convey messages	(Karimipour et al., 2010)
Quantum networking	standard TEU container equipped with error-corrected quantum memories	(Devitt et al., 2014)

A recurrent misconception is that these usages describe one common formalism. They do not. A “quantum parcel” in Edalat’s Interval Quantum Mechanics is an open convex set in state space, whereas a “quantum parcel” in superfluid turbulence is a Lagrangian tracer, and a “quantum parcel” in ship-based networking is a cargo container carrying logical Bell pairs. The phrase therefore functions as a family of domain-specific constructions rather than a unified technical term.

2. Finite-precision state geometry in Interval Quantum Mechanics

In Interval Quantum Mechanics, a quantum parcel is a basic weak$^*$-open set in the state space

$$\mathcal{D}(\mathcal{H})=\{\rho\in\mathcal{T}(\mathcal{H}):\rho\ge0,\;\mathrm{Tr}\,\rho=1\},$$

defined by finitely many open expectation constraints (Edalat, 19 May 2026). Given Hermitian observables $\{A_i\}_{i=1}^k$ and open intervals $(a_i^-,a_i^+)$, the associated parcel is

$$P\bigl(\{A_i,(a^-_i,a^+_i)\}\bigr) =\Bigl\{\rho\in\mathcal{D}(\mathcal{H}): a_i^-<\mathrm{Tr}(A_i\,\rho)<a_i^+\;\forall i\Bigr\}.$$

The physical interpretation is explicit: finite-resolution experiments do not determine a point state but rather the set of all density matrices compatible with the measured macroscopic data.

The framework assigns a Hilbert–Schmidt volume to parcels. When a parcel is defined by orthonormal observables $H_1,\dots,H_m$ with interval lengths $\Delta_i=a_i^+-a_i^-$, the volume is

$$\mathrm{Vol}\bigl(P(\{H_i,(a^-_i,a^+_i)\})\bigr)=\prod_{i=1}^m\Delta_i,$$

and the geometric information is

$(p_c(t),x_c(t))$0

Unitary Schrödinger evolution lifts to a deterministic flow on parcels via

$(p_c(t),x_c(t))$1

with preservation of openness, convexity, compact closure, and Hilbert–Schmidt volume (Edalat, 19 May 2026).

Finite-precision measurement is modeled by a fuzzy POVM. Starting from sharp projectors $(p_c(t),x_c(t))$2, one defines

$(p_c(t),x_c(t))$3

and updates a parcel upon outcome $(p_c(t),x_c(t))$4 by

$(p_c(t),x_c(t))$5

Under the stated positivity conditions, $(p_c(t),x_c(t))$6, so measurement strictly increases geometric information (Edalat, 19 May 2026). Edalat’s double-parcel construction introduces a second, impossible set $(p_c(t),x_c(t))$7, with information ratio

$(p_c(t),x_c(t))$8

which increases monotonically under fuzzy measurement.

A second IQM paper extends parcels to nonequilibrium statistical mechanics. For a parcel $(p_c(t),x_c(t))$9 inside an energy shell, the effective dimension is defined as

$^*$0

and if every state in the parcel has large effective dimension, then for most late times the expectation interval of any bounded observable concentrates around the microcanonical value (Edalat, 30 May 2026). For double parcels $^*$1 separated by a conserved quantity $^*$2, both components thermalize on non-conserved observables, the separation by $^*$3 is preserved exactly, and fuzzy-measurement updates remain valid. In this usage, quantum parcels are neither particles nor messages; they are finite-precision epistemic state regions.

3. Dynamical parcels in walks, phase-space flows, and superfluids

A distinct dynamical usage appears in one-dimensional discrete-time quantum walks. Vieira, Rigolin and Amorim encode a qubit in the spin degree of freedom and localize it around $^*$4 with a real Gaussian envelope of width $^*$5,

$^*$6

with $^*$7 (Vieira et al., 2021). The evolution uses only two local $^*$8 unitaries: the Hadamard coin

$^*$9

and the reflection coin

$$\mathcal{D}(\mathcal{H})=\{\rho\in\mathcal{T}(\mathcal{H}):\rho\ge0,\;\mathrm{Tr}\,\rho=1\},$$0

In a translation-invariant Hadamard walk the eigenphases satisfy $$\mathcal{D}(\mathcal{H})=\{\rho\in\mathcal{T}(\mathcal{H}):\rho\ge0,\;\mathrm{Tr}\,\rho=1\},$$1, with group velocity

$$\mathcal{D}(\mathcal{H})=\{\rho\in\mathcal{T}(\mathcal{H}):\rho\ge0,\;\mathrm{Tr}\,\rho=1\},$$2

so near $$\mathcal{D}(\mathcal{H})=\{\rho\in\mathcal{T}(\mathcal{H}):\rho\ge0,\;\mathrm{Tr}\,\rho=1\},$$3 the wave packet splits into two counterpropagating Gaussians moving at $$\mathcal{D}(\mathcal{H})=\{\rho\in\mathcal{T}(\mathcal{H}):\rho\ge0,\;\mathrm{Tr}\,\rho=1\},$$4. By placing $$\mathcal{D}(\mathcal{H})=\{\rho\in\mathcal{T}(\mathcal{H}):\rho\ge0,\;\mathrm{Tr}\,\rho=1\},$$5 at sites $$\mathcal{D}(\mathcal{H})=\{\rho\in\mathcal{T}(\mathcal{H}):\rho\ge0,\;\mathrm{Tr}\,\rho=1\},$$6 as reflecting “fence posts” and switching them at prescribed times, the protocol achieves long-lived confinement and long-distance release and recapture. For $$\mathcal{D}(\mathcal{H})=\{\rho\in\mathcal{T}(\mathcal{H}):\rho\ge0,\;\mathrm{Tr}\,\rho=1\},$$7 and $$\mathcal{D}(\mathcal{H})=\{\rho\in\mathcal{T}(\mathcal{H}):\rho\ge0,\;\mathrm{Tr}\,\rho=1\},$$8, the trapping time is $$\mathcal{D}(\mathcal{H})=\{\rho\in\mathcal{T}(\mathcal{H}):\rho\ge0,\;\mathrm{Tr}\,\rho=1\},$$9 steps with $\{A_i\}_{i=1}^k$0; single-shot transport over a few hundred sites gives $\{A_i\}_{i=1}^k$1 after $\{A_i\}_{i=1}^k$2 steps; multiple-station transport reaches $\{A_i\}_{i=1}^k$3 sites in $\{A_i\}_{i=1}^k$4 steps with $\{A_i\}_{i=1}^k$5 (Vieira et al., 2021). Here the “parcel” is a nearly dispersionless Gaussian qubit packet.

In the Husimi representation, the term denotes elementary parcels of probability fluid. The Husimi function

$\{A_i\}_{i=1}^k$6

is a Gaussian convolution of the Wigner function and is nonnegative and normalized, so it can be treated as a genuine phase-space density (Zhdanov et al., 2021). A pure state has Husimi representation

$\{A_i\}_{i=1}^k$7

and each small Gaussian packet is an “elementary parcel” parameterized by $\{A_i\}_{i=1}^k$8. The paper proves that the motions of elementary parcels of both classical and quantum Husimi fluid obey the Hamilton variational principle. It also identifies a Skodje flux gauge freedom: different gauge choices can dramatically alter parcel trajectories while leaving the continuity equation for the full Husimi density invariant (Zhdanov et al., 2021).

Superfluid turbulence provides yet another meaning. In the vortex-filament model, a superfluid parcel is a passive tracer satisfying

$\{A_i\}_{i=1}^k$9

where $(a_i^-,a_i^+)$0 is the Biot–Savart velocity field generated by the vortex tangle (Tang et al., 1 Apr 2025). The study distinguishes ultra-quantum turbulence (UQT), with a random tangle and no large-scale polarization, from quasiclassical turbulence (QCT), with locally bundled vortices and a Kolmogorov-like inertial range. For single-body diffusion, UQT parcels exhibit $(a_i^-,a_i^+)$1 with $(a_i^-,a_i^+)$2 for $(a_i^-,a_i^+)$3, whereas QCT parcels are ballistic with $(a_i^-,a_i^+)$4 on the same timescale. For two-body dispersion, UQT parcels satisfy $(a_i^-,a_i^+)$5 with $(a_i^-,a_i^+)$6, while QCT parcels follow the Richardson–Obukhov form $(a_i^-,a_i^+)$7 with $(a_i^-,a_i^+)$8 (Tang et al., 1 Apr 2025). In this setting parcels are not wave packets but inertialess tracers used to probe transport in inviscid quantum fluids.

4. Parcels as transportable entanglement resources

In quantum communication, Marvian and Karimipour introduced the quantum carrier, also described as a quantum parcel, as a fixed entangled state shared once at the start of a protocol and then reused to convey classical or quantum messages (Karimipour et al., 2010). In the two-party case the carrier is the Bell pair

$(a_i^-,a_i^+)$9

The sender uploads the message by local CNOT operations, the receiver downloads it by local unitaries on the carrier share and the traveling qudits, and after download the carrier is restored to its original form. During transit the reduced state of the transmitted system is maximally mixed, and for threshold schemes the access structure is implemented with CSS/polynomial codewords. In this usage, a parcel is a reusable entanglement scaffold.

A much more literal transport model appears in “High-speed quantum networking by ship,” where a quantum parcel is a standard Twenty-foot Equivalent Unit container carrying error-corrected quantum memories (Devitt et al., 2014). One cubic meter is devoted to the memory bank and the remaining $$P\bigl(\{A_i,(a^-_i,a^+_i)\}\bigr) =\Bigl\{\rho\in\mathcal{D}(\mathcal{H}): a_i^-<\mathrm{Tr}(A_i\,\rho)<a_i^+\;\forall i\Bigr\}.$$0 to power, refrigeration, shielding, and classical control. A single logical qubit is encoded in a planar code with $$P\bigl(\{A_i,(a^-_i,a^+_i)\}\bigr) =\Bigl\{\rho\in\mathcal{D}(\mathcal{H}): a_i^-<\mathrm{Tr}(A_i\,\rho)<a_i^+\;\forall i\Bigr\}.$$1 physical qubits; with $$P\bigl(\{A_i,(a^-_i,a^+_i)\}\bigr) =\Bigl\{\rho\in\mathcal{D}(\mathcal{H}): a_i^-<\mathrm{Tr}(A_i\,\rho)<a_i^+\;\forall i\Bigr\}.$$2, $$P\bigl(\{A_i,(a^-_i,a^+_i)\}\bigr) =\Bigl\{\rho\in\mathcal{D}(\mathcal{H}): a_i^-<\mathrm{Tr}(A_i\,\rho)<a_i^+\;\forall i\Bigr\}.$$3, and $$P\bigl(\{A_i,(a^-_i,a^+_i)\}\bigr) =\Bigl\{\rho\in\mathcal{D}(\mathcal{H}): a_i^-<\mathrm{Tr}(A_i\,\rho)<a_i^+\;\forall i\Bigr\}.$$4, the example gives $$P\bigl(\{A_i,(a^-_i,a^+_i)\}\bigr) =\Bigl\{\rho\in\mathcal{D}(\mathcal{H}): a_i^-<\mathrm{Tr}(A_i\,\rho)<a_i^+\;\forall i\Bigr\}.$$5, $$P\bigl(\{A_i,(a^-_i,a^+_i)\}\bigr) =\Bigl\{\rho\in\mathcal{D}(\mathcal{H}): a_i^-<\mathrm{Tr}(A_i\,\rho)<a_i^+\;\forall i\Bigr\}.$$6, and $$P\bigl(\{A_i,(a^-_i,a^+_i)\}\bigr) =\Bigl\{\rho\in\mathcal{D}(\mathcal{H}): a_i^-<\mathrm{Tr}(A_i\,\rho)<a_i^+\;\forall i\Bigr\}.$$7 days. Logical Bell-pair creation uses lattice surgery with

$$P\bigl(\{A_i,(a^-_i,a^+_i)\}\bigr) =\Bigl\{\rho\in\mathcal{D}(\mathcal{H}): a_i^-<\mathrm{Tr}(A_i\,\rho)<a_i^+\;\forall i\Bigr\}.$$8

and the link fidelity after transport is written as

$$P\bigl(\{A_i,(a^-_i,a^+_i)\}\bigr) =\Bigl\{\rho\in\mathcal{D}(\mathcal{H}): a_i^-<\mathrm{Tr}(A_i\,\rho)<a_i^+\;\forall i\Bigr\}.$$9

For the NV$H_1,\dots,H_m$0 example, $H_1,\dots,H_m$1 Ebits per container, $H_1,\dots,H_m$2, $H_1,\dots,H_m$3, and $H_1,\dots,H_m$4 Ebit/s. The proposal explicitly contrasts containerized memories with undersea repeater deployment.

The EPR-packet literature uses “parcel” for buffered entangled resources in memory nodes. A Markov-chain model represents a node storing up to $H_1,\dots,H_m$5 entangled pairs and performing up to $H_1,\dots,H_m$6 rounds of DEJMPS distillation per communication cycle (Valentini et al., 15 Apr 2026). The state is $H_1,\dots,H_m$7, with $H_1,\dots,H_m$8, and the distillation success probability is

$H_1,\dots,H_m$9

The paper derives expressions for fidelity decay, throughput,

$\Delta_i=a_i^+-a_i^-$0

outage probability, and memory-dimensioning rules for a target rate and fidelity. Here the parcel is not the memory hardware itself but an EPR packet carried and managed by the memory architecture.

A common misconception is to conflate these three notions. The quantum carrier is a reusable entangled state, the ship-borne quantum parcel is a cargo container with fault-tolerant memories, and the EPR packet is a buffered entanglement unit in a memory-sizing model.

5. Parcel delivery as a target for quantum optimization

A separate research line uses quantum computing to optimize parcel-delivery operations. In these papers “parcel” refers to packages, customers, schedules, trucks, or drones rather than a physical quantum object.

Q4RPD, “Quantum for Real Package Delivery,” formulates realistic last-mile routing with D-Wave’s hybrid CQM stack. The 2024 paper defines the 2-Dimensional and Heterogeneous Package Delivery with Priorities problem, with two-dimensional capacity, top-priority delivery deadlines, route-duration limits, heterogeneous fleets, and business preferences favoring owned vehicles (Osaba et al., 2024). Because the full model is too large for current hardware, the method decomposes the problem into Single Routing Problems and submits each constrained quadratic model to LeapCQMHybrid, which combines classical heuristics with quantum-annealer subroutines on an Advantage_system6.4 QPU. Across six instances with 14–29 deliveries, all constraints were met; in non-TP cases Q4RPD matched the OR-Tools TSP optimum, and in TP-rich cases the gap was at most $\Delta_i=a_i^+-a_i^-$1, attributed to enforcing the deadline constraint.

The 2025 extension of Q4RPD adds simultaneous pickup and delivery, time windows, and mobility restrictions by vehicle type (Osaba et al., 2 Apr 2025). The model uses binary variables $\Delta_i=a_i^+-a_i^-$2 indicating whether order $\Delta_i=a_i^+-a_i^-$3 is visited in route position $\Delta_i=a_i^+-a_i^-$4, capacity constraints for weight and volume at every route position, linear-inequality time-window constraints, and either filtering or explicit constraints for vehicle accessibility. The implementation again uses LeapCQMHybrid, now with Advantage_system7.1, a $\Delta_i=a_i^+-a_i^-$5 wall-clock time per route, and seven instances up to 24 nodes. All seven returned feasible, capacity- and window-respecting routes within $\Delta_i=a_i^+-a_i^-$6 per route, with total wall time under $\Delta_i=a_i^+-a_i^-$7 even on the largest instances (Osaba et al., 2 Apr 2025).

A gate-model reinforcement-learning approach appears in “Quantum-Efficient Reinforcement Learning Solutions for Last-Mile On-Demand Delivery,” which models the task as a Capacitated Pickup and Delivery Problem with Time Windows (Moosavi et al., 7 Aug 2025). The objective is

$\Delta_i=a_i^+-a_i^-$8

and the policy network is replaced by a parametrized quantum circuit with $\Delta_i=a_i^+-a_i^-$9 qubits, a Hadamard superposition layer, load and time-window encoding rotations, controlled-$$\mathrm{Vol}\bigl(P(\{H_i,(a^-_i,a^+_i)\})\bigr)=\prod_{i=1}^m\Delta_i,$$0 gates on pickup–delivery pairs, distance-weighted Ising-$$\mathrm{Vol}\bigl(P(\{H_i,(a^-_i,a^+_i)\})\bigr)=\prod_{i=1}^m\Delta_i,$$1 gates, and trainable $$\mathrm{Vol}\bigl(P(\{H_i,(a^-_i,a^+_i)\})\bigr)=\prod_{i=1}^m\Delta_i,$$2–$$\mathrm{Vol}\bigl(P(\{H_i,(a^-_i,a^+_i)\})\bigr)=\prod_{i=1}^m\Delta_i,$$3 layers. With $$\mathrm{Vol}\bigl(P(\{H_i,(a^-_i,a^+_i)\})\bigr)=\prod_{i=1}^m\Delta_i,$$4 nodes, $$\mathrm{Vol}\bigl(P(\{H_i,(a^-_i,a^+_i)\})\bigr)=\prod_{i=1}^m\Delta_i,$$5 vehicles, 10 000 random instances, and learning rate $$\mathrm{Vol}\bigl(P(\{H_i,(a^-_i,a^+_i)\})\bigr)=\prod_{i=1}^m\Delta_i,$$6, the paper reports that PQC-PPO with $$\mathrm{Vol}\bigl(P(\{H_i,(a^-_i,a^+_i)\})\bigr)=\prod_{i=1}^m\Delta_i,$$7 converges after $$\mathrm{Vol}\bigl(P(\{H_i,(a^-_i,a^+_i)\})\bigr)=\prod_{i=1}^m\Delta_i,$$8 epochs, whereas $$\mathrm{Vol}\bigl(P(\{H_i,(a^-_i,a^+_i)\})\bigr)=\prod_{i=1}^m\Delta_i,$$9 needs $(p_c(t),x_c(t))$00 epochs; PQC-PPO achieves the lowest median cost and the tightest interquartile range among DDQN, classical PPO, and QSVT-based baselines (Moosavi et al., 7 Aug 2025).

Drone scheduling and packing have been mapped to independent-set covering on a neutral-atom platform. Tarquini et al. define a scheduling graph $(p_c(t),x_c(t))$01 in which deliveries conflict when their time intervals overlap, and cover all deliveries with the minimum number of battery-feasible independent sets (Tarquini et al., 17 Feb 2026). The neutral-atom device implements the independent-set sampler through the Rydberg Hamiltonian

$(p_c(t),x_c(t))$02

The hybrid workflow performs QPU sampling, classical repair of independence and battery feasibility, and a final ILP covering step. Emulation up to $(p_c(t),x_c(t))$03 deliveries yields approximation ratio $(p_c(t),x_c(t))$04 within a few percent of 1 up to $(p_c(t),x_c(t))$05, and hardware experiments on Pasqal’s Fresnel QPU report $(p_c(t),x_c(t))$06 at $(p_c(t),x_c(t))$07, $(p_c(t),x_c(t))$08 at $(p_c(t),x_c(t))$09, $(p_c(t),x_c(t))$10 at $(p_c(t),x_c(t))$11, and $(p_c(t),x_c(t))$12 at $(p_c(t),x_c(t))$13 (Tarquini et al., 17 Feb 2026).

A gate-based exact-search direction is developed for CVRPTW in the Grover framework (Meghazi et al., 18 May 2026). The model stores a giant tour $(p_c(t),x_c(t))$14, binary split variables $(p_c(t),x_c(t))$15, cumulative loads $(p_c(t),x_c(t))$16, and arrival times $(p_c(t),x_c(t))$17, and uses a reversible oracle to check all-different, capacity, time-window, and cost-threshold constraints. The encoding adds only a linear number of decision qubits beyond standard TSP formulations, while the total space complexity is dominated by $(p_c(t),x_c(t))$18 ancillas for all-different checks. The paper explicitly states that even $(p_c(t),x_c(t))$19 customers requires on the order of 150 qubits and $(p_c(t),x_c(t))$20 MCX gates per Grover iterate, placing the method beyond current NISQ hardware (Meghazi et al., 18 May 2026).

6. Related extensions, metaphorical uses, and conceptual boundaries

The term also appears in more speculative or auxiliary senses. In loop quantum gravity, Achour and collaborators introduce “puncels,” a contraction of puncture and parcel, in the fluid approximation of a quantum spherical black hole (Heidmann et al., 2016). A puncel is the large-spin intertwiner associated with one tetrahedron in a triangulation of the black-hole interior, with horizon and internal labels scaling as $(p_c(t),x_c(t))$21. The semi-classical Euclidean volume operator on a puncel yields fluctuation scales

$(p_c(t),x_c(t))$22

and the largest scale is then used to motivate an effective temperature proportional to $(p_c(t),x_c(t))$23, reproducing Hawking scaling (Heidmann et al., 2016). The construction is explicitly approximate: it assumes the fluid regime, Euclidean continuation, and diagonal dominance of the volume operator.

A different metaphor appears in noncommutative-geometry-inspired work connecting diffusion and Schrödinger dynamics. “Parcels of Universe” argues that volume quantization can be interpreted as a random mosaic of elementary blocks of volume $(p_c(t),x_c(t))$24 and $(p_c(t),x_c(t))$25, and uses this picture to motivate a complex stochastic process behind the Schrödinger equation (Frasca et al., 2018). The route runs from the Fourier equation

$(p_c(t),x_c(t))$26

to the free-particle Schrödinger equation via the Wick rotation $(p_c(t),x_c(t))$27 and $(p_c(t),x_c(t))$28. The paper then links this formal relation to quantized space parcels in the Connes–Chamseddine–Mukhanov framework (Frasca et al., 2018). This is conceptually distinct from both IQM parcels and Husimi parcels.

Two boundaries are therefore important. First, not every “quantum parcel” is a subsystem or a carrier of transferable information; in IQM it is a finite-precision open set, and in superfluid turbulence it is a passive tracer. Second, not every parcel-related paper is about a novel quantum state concept; many are optimization papers about package delivery. The shared vocabulary should not obscure the fact that these works sit in different mathematical categories: convex state-space geometry, wave-packet engineering, tracer kinematics, topological quantum networking, quantum-memory queueing, and quantum optimization.

Taken together, the literature shows that “quantum parcels” is a productive but heterogeneous label. It names state-space regions in finite-precision quantum mechanics, localized dynamical objects in walks and fluids, entanglement-bearing communication resources, and application domains in which quantum hardware or algorithms are used to move physical parcels more efficiently.