[Paper Review] Towards Prototyping Driverless Vehicle Behaviors, City Design, and Policies Simultaneously

Hauke Sandhaus, Wendy Ju+1 more|arXiv (Cornell University)|Apr 13, 2023

Transportation and Mobility Innovations4 citations

TL;DR

This paper proposes a simultaneous prototyping framework for autonomous vehicles (AVs), urban design, and policy through two integrated approaches: low-fidelity integrative City-AV-Policy Simulation (iCAPS) and participatory design optimization. By combining digital twins, machine learning models, and stakeholder-driven design workflows, it enables collaborative, computationally feasible, and ethically informed co-design of AV-integrated cities.

ABSTRACT

Autonomous Vehicles (AVs) can potentially improve urban living by reducing accidents, increasing transportation accessibility and equity, and decreasing emissions. Realizing these promises requires the innovations of AV driving behaviors, city plans and infrastructure, and traffic and transportation policies to join forces. However, the complex interdependencies among AV, city, and policy design issues can hinder their innovation. We argue the path towards better AV cities is not a process of matching city designs and policies with AVs' technological innovations, but a process of iterative prototyping of all three simultaneously: Innovations can happen step-wise as the knot of AV, city, and policy design loosens and tightens, unwinds and reties. In this paper, we ask: How can innovators innovate AVs, city environments, and policies simultaneously and productively toward better AV cities? The paper has two parts. First, we map out the interconnections among the many AV, city, and policy design decisions, based on a literature review spanning HCI/HRI, transportation science, urban studies, law and policy, operations research, economy, and philosophy. This map can help innovators identify design constraints and opportunities across the traditional AV/city/policy design disciplinary bounds. Second, we review the respective methods for AV, city, and policy design, and identify key barriers in combining them: (1) Organizational barriers to AV-city-policy design collaboration, (2) computational barriers to multi-granularity AV-city-policy simulation, and (3) different assumptions and goals in joint AV-city-policy optimization. We discuss two broad approaches that can potentially address these challenges, namely, "low-fidelity integrative City-AV-Policy Simulation (iCAPS)" and "participatory design optimization".

Motivation & Objective

• Address the complex interdependence between AV behavior design, urban infrastructure, and policy frameworks, which collectively form a tightly coupled 'knot' of design challenges.
• Overcome the lack of coordinated innovation across AV, city, and policy domains by proposing a shared, iterative prototyping process.
• Enable cross-disciplinary collaboration among AV engineers, urban planners, and policymakers through simulation and participatory methods.
• Identify and address key barriers—organizational, computational, and goal-oriented—that hinder simultaneous AV-city-policy innovation.
• Catalyze a principled, collaborative design process that aligns technological development with societal values and urban livability goals.

Proposed method

• Develop a conceptual map of interdependencies between AV behaviors, urban design, and policy using a cross-disciplinary literature review across HCI, transportation science, urban studies, law, and policy.
• Propose 'low-fidelity integrative City-AV-Policy Simulation' (iCAPS), leveraging digital twins for high-fidelity visualizations and modifiable urban environments, combined with machine learning models to simulate road user behaviors.
• Integrate machine learning models that predict behavioral outcomes (e.g., safety, traffic flow) under different AV policy changes, such as right-of-way rules.
• Implement a participatory design optimization workflow that starts with stakeholder workshops to define success metrics and design priorities.
• Use Bayesian optimization within the defined problem space to computationally identify AV, city, and policy designs that best reflect stakeholder preferences and trade-offs.
• Create a feedback loop where stakeholders evaluate tentative optimal designs before deployment, ensuring legitimacy and real-world relevance.

Experimental results

Research questions

• RQ1 How can AV, urban design, and policy decisions be simultaneously prototyped to avoid misalignment and ensure mutual compatibility?
• RQ2 What are the key organizational, computational, and goal-related barriers to integrating AV, city, and policy design in a shared innovation process?
• RQ3 How can low-fidelity simulation tools effectively visualize and predict the impacts of AV behavior, urban form, and policy changes on traffic and urban life?
• RQ4 What role can participatory design play in aligning diverse stakeholder values with technical and urban design outcomes in AV integration?
• RQ5 How can machine learning models and digital twins be combined to enable real-time, on-demand simulation of AV-policy-urban design interactions under computational constraints?

Key findings

• The interdependence among AV behavior, urban design, and policy forms a tightly coupled 'knot' that cannot be untangled by optimizing one domain in isolation.
• Low-fidelity integrative simulation (iCAPS) enables shared, real-time prototyping by combining digital twins with ML-based behavior prediction, reducing computational and cognitive load.
• Participatory design optimization successfully frames the design problem around stakeholder-defined metrics and priorities, enabling data-driven, value-sensitive design choices.
• Computational efficiency in simulation is achievable through strategic abstraction—visualizations and predictions need only be precise where relevant to the current design proposal.
• Ethical and practical data-sharing across AV companies, urban agencies, and regulators is a prerequisite for scalable, trustworthy iCAPS platforms.
• The integration of participatory methods with optimization techniques leads to more legitimate, stakeholder-validated design outcomes than top-down or purely technical approaches.