Field Validation of a Portable AI Smartphone Microplastic-Nanoplastic Test in Coastal Waters

This paper is also available at:

https://doi.org/10.5281/zenodo.19490782 This field validation study evaluated the robustness and real-world deployability of a portable smartphone based zero-shear optical interaction assay for detection of microplastics and nanoplastics (MP–NP) in coastal water at Crissy Field East Beach, San Francisco Bay — a representative estuarine environment with elevated salinity and known microplastic presence.

Samples were collected directly from the shoreline and analyzed under fully uncontrolled field conditions, including natural sunlight with variable intensity and shadowing, partial body and hand occlusion, an uneven wooden surface, and a hand-drawn marker grid, without the use of a photobox or standardized imaging setup.

Zero-Shear Protocol

A zero-shear protocol was used, in which the reagent is introduced directly into the water sample without mixing or agitation. Mechanical mixing (even gentle stirring by hand) can introduce variable shear forces that alter particle interaction dynamics and produce inconsistent optical patterns. By eliminating post-addition disturbance, the protocol preserves intrinsic interaction-driven spatial organization, resulting in more reproducible optical signals.

Operationally, the workflow is intentionally simple:

(1) collect the water sample,

(2) introduce the reagent and allow the system to evolve without disturbance, and

(3) capture an image without contacting or perturbing the sample.

This minimal-intervention approach supports both ease of use and consistent data acquisition for downstream computer vision analysis.

Treated samples produced reproducible, time-dependent optical patterns using AI and computer vision.  The optical patterns were characterized by progression from diffuse haze at early time points to distinct radial clearing and recognizable structure formation over time. These interaction-driven spatial phenotypes were clearly distinguishable from control samples, which exhibited only passive sedimentation and maintained sharp background visibility. Independent replicate samples collected in close proximity demonstrated consistent kinetic progression and spatial organization.

Environmental water samples are known to have both microplastics and nanoplastics within them.  The observed optical signatures are consistent with interaction regimes expected in mixed microplastic–nanoplastic water samples as observed in standard laboratory conditions, indicating that detection behavior persists in authentic high-ionic-strength coastal water.

Importantly, assay performance remained stable despite significant deviations from controlled laboratory conditions, demonstrating that the diagnostic signal arises from intrinsic spatial organization of particulate interactions rather than dependence on lighting conditions, color fidelity, or specialized imaging hardware.

These results establish the suitability of the assay for scalable field deployment in real-world coastal and marine environments and support its use in decentralized environmental monitoring applications, including research, policy, and citizen science contexts.

Keywords

microplastics, nanoplastics, optical assay, field validation, coastal water, San Francisco Bay, zero-shear, environmental monitoring, portable diagnostics, EcoExposure