Every number on this page is computed by the Volume II Chapter 6 notebook's own code; the instance is deterministic, and its random feasible plans use the lab's seed 26206. Your notes stay on this page until you leave it; copy the memo before you go.
The sale of part of Industrial Systems is on the board's agenda. The chief financial officer wants the buyer's extra $20 million; the chief human resources officer must plan the rebuild if the engineering team goes; the transformation office worries about the digital build's start date. You support the team preparing the board's decision.
Use AXIOM to solve Meridian's capability problem by the maximum principle and by transcription, read the price of a point, and test the answer against the declarations it rests on.
Worked Example 6.1: the state is capability c; the decision is the rate of hiring and training u, at $4 million a point and at most 2.5 points a quarter; ċ = u − δc with δ = 0.101 a year. The digital build runs from the first moment capability reaches 55, and from then on capability must stay at or above 55; every year it waits costs $32 million. Discounted at ρ = ln 1.1 over an open horizon.
In the state simulator, run Volume I's schedules and the race-and-rent plan; compare capability paths and the build's start date.
Present values at 10 percent over an open horizon, $ million. Proceeds: $92 million with the team kept, $112 million with it gone.
In the Hamiltonian visualization, plot the Hamiltonian against the hiring rate and the switching function m − 4 along each plan.
H = −4u − 32·1{c < 55} + m(u − δc). The maximum condition reads off (m − 4)u: hire at the cap if m > 4, nothing if m < 4, any rate if m = 4 (Theorem 6.2).
In the costate analysis, compute the price of a point along each plan and check it by finite differences.
In optimal policy computation, solve by shooting, and by direct transcription handed to the AXIOM-04 solvers; cross-audit the two.
Run the state and the costate forward with the decision the maximum condition picks, and correct the starting price by the secant rule until the plan ends where it must: for the race, the price on arrival equals the point's cost plus the delay it saves; for the drift, the price reaches the cost as capability reaches the floor.
In strategy comparison, vary attrition, the hiring cap, the value of a quarter of delay, and the floor's margin, and record the value of the team.
Prices that grow at the hurdle rate plus the attrition rate off the floor and sit at the cost of a point on it (Theorem 6.1); decisions at the cap or at zero according to the sign of the switching function (Theorem 6.2, Proposition 6.2); a floor priced at the user cost (Proposition 6.3); and a certificate that holds where the problem is concave and must be argued directly where it is not (Theorem 6.3).
The value of the team, the rebuild plan if it goes, the price of the floor and of a margin over it, and the declarations the answer depends on.