I recently listened to an interview with Tomas Haerdin on planning which
is an economic production system that focuses on creation of physical
items. It was most obviously a strategy employed by the USSR, but in the
interview he also goes over how it was used by the Allende government in
Chile under Project Cybersyn.

Project Cybersyn introduced  the interesting concept of the discussion,
cybernetic planning. It was different than the Soviet Gosplan system in
that it was a "bottom-up" planned economy, with aspirations of
democratic worker control.

Haerdin's describes cybernetic planning as a systems theory with
feedback loops. From a technical level, it models both a distributed
system of production, as well as solvers for systems of equations.

After being out of school for some time the the linear programming went
over my head. But he estimates that 23 billion production methods
(equations) can be solved on modern hardware. These models, as well as
representing production firms, could then be represented as a graph. 

Representing firms in the system, software could enable workers to
propose a new product for production. The proposal would be modeled on
the cost of inputs and outputs, and then the planning system would be
able to direct resources from existing goals to newly proposed and
approved products.

This discussion also reminds me of "The People's Republic of Walmart",
and the solving of such large systems of equations piqued my interest.