Table 2 Network costs and performance. Comparison of key observables of the planned networks by batched direct optimization (DO) and by backward percolation (BP) with the dynamic-demand evaluation function \(Q_\textrm{dyn}\). After year 3, both planned networks have similar total length and maintenance costs (MC). Though the construction costs (CC) of BP is 13% smaller compared to DO, it achieves 9% larger net present value (NPV) due to larger travel time benefits (TB). After year 10, both networks still have a similar total length with almost identical construction costs. Yet, the network planned by BP has 5% higher maintenance costs, reducing the available budget for future investments. Also, DO already achieves a 11% higher net present value over the BP despite lower welfare gains due to its advantage in health benefits (HB). After year 30, the batched direct optimization no longer adds additional segments to the network and maintains a net present value approximately 8 to 10% higher than the backward percolation approach. See Supplementary Note 4, Fig. S3 for the cost and performance of all five approaches.

From: Robust design of bicycle infrastructure networks

 

Year 3

Year 10

Year 30

Year 50

\(G_\textrm{full}\)

 

DO

BP

DO

BP

DO

BP

DO

BP

 

Total length built [km]

338

350

693

705

1331

1388

1331

1876

1876

Relative length built [percent]

18.01

18.65

36.94

37.58

70.92

73.98

70.92

100.00

100.00

\(\textrm{CC}_\textrm{rel}\) [percent]

6.86

6.01

20.58

20.56

54.34

57.41

54.34

100.00

100.00

\(\textrm{NPV}\) [m EUR]

41.56

45.35

584.75

522.06

2321.80

2074.91

3585.40

3284.11

-

\(\textrm{TB}\) [m EUR]

14.66

17.82

141.64

149.49

575.12

562.41

886.72

873.28

-

\(\textrm{HB}\) [m EUR]

43.00

41.87

498.14

427.20

1912.15

1679.53

2929.26

2683.81

-

\(\textrm{CC}\) [m EUR]

15.18

13.28

45.50

45.46

120.14

126.92

120.14

221.10

221.10

\(\textrm{MC}\) [m EUR]

0.72

0.73

3.08

3.24

9.30

9.78

9.30

18.20

18.20