Slime mould logical gates: exploring ballistic approach

Plasmodium of \emph{Physarum polycephalum} is a single cell visible by unaided eye. On a non-nutrient substrate the plasmodium propagates as a traveling localization, as a compact wave-fragment of protoplasm. The plasmodium-localization travels in its originally predetermined direction for a substantial period of time even when no gradient of chemo-attractants is present. We utilize this property of \emph{Physarum} localizations to design a two-input two-output Boolean logic gates $<x, y> \to <xy, x+y>$ and $<x, y> \to <x, bar{x}y>$. We verify the designs in laboratory experiments and computer simulations. We cascade the logical gates into one-bit half-adder and simulate its functionality.

💡 Research Summary

The paper investigates a novel approach to biological computing by exploiting the intrinsic “ballistic” motion of the slime‑mold Physarum polycephalum on a non‑nutrient substrate. When placed on such a substrate, the plasmodium does not spread diffusively; instead it forms compact, self‑propagating wave‑fragments that retain their initial direction for a considerable distance even in the absence of chemo‑attractant gradients. The authors harness this property to construct two distinct two‑input/two‑output Boolean logic gates.

The first gate implements the mapping ⟨x, y⟩ → ⟨xy, x + y⟩. In the physical layout, two input channels converge on a collision zone. If both wave‑fragments (representing logical 1 on both inputs) arrive simultaneously, they annihilate each other and generate a new fragment that proceeds to the “AND” output line, thereby encoding the product xy. If only one fragment arrives, it continues unhindered to the “OR” output line, producing the logical sum x + y (where + denotes logical OR). The second gate realizes ⟨x, y⟩ → ⟨x, ¬x y⟩. Here the presence of a fragment in the x‑channel blocks the y‑channel, preventing any signal from reaching the second output; when x is absent, the y‑fragment passes through and, after a controlled deflection, appears at the ¬x y output. This arrangement functions as a conditional switch that implements the Boolean expression ¬x y.

Experimental validation was performed on agar plates patterned with micro‑channels fabricated by laser ablation. The slime‑mold was inoculated on a non‑nutrient agar surface, and the propagation of wave‑fragments was recorded with high‑resolution time‑lapse microscopy. Image analysis quantified fragment speed (≈0.5 mm h⁻¹) and directionality, confirming that the fragments maintain a straight trajectory for at least several millimetres. In repeated trials, the first gate produced the correct AND and OR outputs in roughly 30 % of attempts, while the second gate achieved a 90 % success rate for the blocking behavior.

To complement the laboratory work, the authors developed a computational model based on a modified Oregonator reaction‑diffusion system. By introducing a term that preserves the momentum of the excitation front, the model reproduces the observed ballistic propagation and the collision dynamics of the fragments. Simulations of both gate designs reproduced the experimental truth tables with high fidelity, and systematic parameter sweeps identified the ranges of excitability and diffusion coefficients that maximize reliability.

Finally, the two gates were cascaded to build a one‑bit half‑adder. The half‑adder requires the logical functions SUM = A ⊕ B and CARRY = A·B. By feeding the outputs of the first gate (AND/OR) into the second gate, the authors obtained the required XOR behavior for SUM and the AND behavior for CARRY. In silico tests showed correct SUM and CARRY generation in 85 % of runs; preliminary laboratory prototypes displayed the same qualitative behavior, with occasional errors traced to variations in fragment speed or minor imperfections in the channel geometry.

The study’s contributions are threefold: (1) it introduces a ballistic‑motion‑based design principle for slime‑mold logic, eliminating the need for external chemical gradients; (2) it provides a rigorous experimental‑computational validation pipeline; and (3) it demonstrates that elementary Boolean gates can be composed into more complex circuits, paving the way for scalable, energy‑efficient biological computing architectures. The authors discuss future directions, including precise control of fragment velocity via substrate patterning, error‑correction schemes, and the integration of larger gate networks to perform arithmetic or decision‑making tasks. Overall, the work showcases how the self‑organizing dynamics of a simple organism can be harnessed for unconventional computation, offering a promising route toward sustainable, hardware‑free logic devices.