MAYBE

Initial complexity problem:
1:	T:
		(Comp: ?, Cost: 1)    f0(Ar_0, Ar_1) -> Com_1(f4(0, 99))
		(Comp: ?, Cost: 1)    f4(Ar_0, Ar_1) -> Com_1(f4(Fresh_1, Ar_1)) [ Ar_1 >= Ar_0 + 1 /\ C >= D + 1 ]
		(Comp: ?, Cost: 1)    f4(Ar_0, Ar_1) -> Com_1(f4(Ar_0, Fresh_0)) [ Ar_1 >= Ar_0 + 1 ]
		(Comp: ?, Cost: 1)    f4(Ar_0, Ar_1) -> Com_1(f11(Ar_0, Ar_1)) [ Ar_0 >= Ar_1 ]
		(Comp: 1, Cost: 0)    koat_start(Ar_0, Ar_1) -> Com_1(f0(Ar_0, Ar_1)) [ 0 <= 0 ]
	start location:	koat_start
	leaf cost:	0

Repeatedly propagating knowledge in problem 1 produces the following problem:
2:	T:
		(Comp: 1, Cost: 1)    f0(Ar_0, Ar_1) -> Com_1(f4(0, 99))
		(Comp: ?, Cost: 1)    f4(Ar_0, Ar_1) -> Com_1(f4(Fresh_1, Ar_1)) [ Ar_1 >= Ar_0 + 1 /\ C >= D + 1 ]
		(Comp: ?, Cost: 1)    f4(Ar_0, Ar_1) -> Com_1(f4(Ar_0, Fresh_0)) [ Ar_1 >= Ar_0 + 1 ]
		(Comp: ?, Cost: 1)    f4(Ar_0, Ar_1) -> Com_1(f11(Ar_0, Ar_1)) [ Ar_0 >= Ar_1 ]
		(Comp: 1, Cost: 0)    koat_start(Ar_0, Ar_1) -> Com_1(f0(Ar_0, Ar_1)) [ 0 <= 0 ]
	start location:	koat_start
	leaf cost:	0

A polynomial rank function with
	Pol(f0) = 1
	Pol(f4) = 1
	Pol(f11) = 0
	Pol(koat_start) = 1
orients all transitions weakly and the transition
	f4(Ar_0, Ar_1) -> Com_1(f11(Ar_0, Ar_1)) [ Ar_0 >= Ar_1 ]
strictly and produces the following problem:
3:	T:
		(Comp: 1, Cost: 1)    f0(Ar_0, Ar_1) -> Com_1(f4(0, 99))
		(Comp: ?, Cost: 1)    f4(Ar_0, Ar_1) -> Com_1(f4(Fresh_1, Ar_1)) [ Ar_1 >= Ar_0 + 1 /\ C >= D + 1 ]
		(Comp: ?, Cost: 1)    f4(Ar_0, Ar_1) -> Com_1(f4(Ar_0, Fresh_0)) [ Ar_1 >= Ar_0 + 1 ]
		(Comp: 1, Cost: 1)    f4(Ar_0, Ar_1) -> Com_1(f11(Ar_0, Ar_1)) [ Ar_0 >= Ar_1 ]
		(Comp: 1, Cost: 0)    koat_start(Ar_0, Ar_1) -> Com_1(f0(Ar_0, Ar_1)) [ 0 <= 0 ]
	start location:	koat_start
	leaf cost:	0

Complexity upper bound ?

Time: 0.467 sec (SMT: 0.439 sec)