Module 01 — Forecast
Forecast
Demand-aware growth. Mycelia predicts where load will surge and pre-grows capacity into those lanes before the orders land.
−18% empty milesmodelled, 12-DC grocery network
How it grows
Scattered nodes. One optimal route.
Drop in your distribution centres, lanes, and demand. Mycelia floods the graph with exploratory agents, then lets the strong paths starve the weak ones — the same reinforcement physics Physarum polycephalum uses to map the shortest route between food sources. What emerges is a network, not a guess.
graph · unsolved growing…
01
Seed
Every facility, lane and demand signal becomes a node in the graph.
02
Flood
Millions of agents explore every possible connection in parallel.
03
Reinforce
Lanes carrying real load thicken; dead ends decay and vanish.
04
Resolve
The surviving veins are your optimized route. Re-solved as conditions shift.
Why Mycelia
A slime mold with no brain can rebuild the Tokyo rail map overnight. Your supply chain deserves the same algorithm.
Most routing software optimizes a snapshot and calls it done. Real networks breathe — demand moves, lanes close, fuel spikes. Mycelia treats your supply chain as a living organism: continuously re-growing the path of least resistance, so the network you run tomorrow is never the one you froze last quarter.
The solver, in three modules
Forecast. Reroute. Consolidate.
Module 02 — Reroute
Reroute
Live re-solving. A closed road, a port delay, a weather event — the network re-grows around the disruption in seconds, not planning cycles.
4.2s re-solvemodelled, 9k-node parcel graph
Module 03 — Consolidate
Consolidate
Vein-merging. Mycelia finds shipments that should share a lane and fuses them — fewer trucks, fuller loads, lower cost per drop.
+11% load factormodelled, cold-chain regional
Case studies
Networks we re-grew.
Illustrative, modelled outcomes on synthetic networks. Filter by industry.
grocery Meridian Foods
−18% empty miles
12 DCs, 340 stores
parcel Halcyon Parcel
4.2s live re-solve
9,000-node last-mile graph
cold-chain Boreal Cold
+11% load factor
Temperature-locked lanes
grocery Verdant Grocers
−9% lane cost
Regional fresh network
parcel Drift Logistics
+14% on-time
Surge-week rerouting
cold-chain Polar Route
−21% spoilage risk
Vaccine cold-chain model
grocery Harvest Co-op
−12% empty miles
Producer-to-shelf
parcel Quill Couriers
−16% fleet idle
Dense-metro consolidation
Modelled outcomes
The numbers the network grows toward.
All figures modelled on synthetic networks — illustrative, not client-reported.
−18%
Empty miles removed
modelled, grocery
4.2s
Live re-solve time
modelled, 9k-node graph
+11%
Load-factor lift
modelled, cold-chain
2.3M
Agents per solve
GPU-parallel
It stopped optimizing a snapshot and started growing a network. The lanes nobody would have drawn by hand are the ones carrying the load now.
VP Network PlanningGrocery, modelled
We watched a port closure ripple through and the route just re-grew around it before our planners had opened the dashboard.
Head of OperationsParcel, modelled
Cold-chain is unforgiving — one broken lane spoils a load. Mycelia keeps the temperature-locked veins thick and the rest lean.
Director of Cold LogisticsCold-chain, modelled
The biophysics framing sounded like a metaphor until the solver beat our incumbent on every synthetic graph we threw at it.
Chief Supply-Chain OfficerMixed-mode, modelled
Re-solving in seconds changes what 'planning' even means. We plan continuously now.
Logistics LeadLast-mile, modelled
Grow your network
Bring us your hardest graph.
Send your network — DCs, lanes, demand — and we'll grow it against your incumbent on a synthetic twin. You see the modelled delta before anything touches production.
Founded by
Mycelia was founded by Dr. Lena Varga, a biophysics PhD who spent six years modelling Physarum reinforcement networks, and Marcus Oyelaran, a former Head of Network Operations who ran 400-lane distribution at continental scale. The thesis: the cheapest router in nature has no brain — only feedback.