Robust design of bicycle infrastructure networks.

Area of Science:

• Transport science
• Network science
• Operations research

Background:

• Effective bicycle networks are crucial for promoting active mobility.
• Expanding these networks presents significant planning challenges due to complex interactions between infrastructure and usage.
• Factors like network spillover effects and mode choice substitutions complicate long-term transport planning.

Purpose of the Study:

• To compare direct optimization methods with network science-based heuristics for planning bicycle network expansions.
• To evaluate the trade-offs between solution quality, complexity, and robustness to uncertainty in transport planning.

Main Methods:

• Direct welfare optimization using operations research techniques.
• Inverse network percolation as a conceptual heuristic approach.
• Comparative analysis of planning approaches for cycle superhighway extensions in Copenhagen.

Main Results:

• More complex optimization models provided only marginal improvements in overall welfare compared to simpler methods.
• Increased planning complexity heightened vulnerability to input uncertainty, illustrating the bias-variance tradeoff.
• The benefits of advanced planning methods were diminished by long-term planning uncertainties.

Conclusions:

• While complex optimization can yield slightly better results, simpler network-based methods offer greater robustness and ease of implementation for bicycle infrastructure planning.
• Robustness and practical implementation considerations may favor straightforward network approaches over complex optimization for long-term transport planning.
• Despite planning complexities, bicycle infrastructure development offers high benefit-cost ratios, underscoring its importance.