Designing Shared Mobility for Public Trust

Overview

Shared mobility, also known as micromobility, now operates as public infrastructure. Shared electric vehicles occupy footpaths, streets, and civic space alongside pedestrians, cyclists, and cars. In this environment, technology decisions directly affect safety, privacy, accessibility, and public confidence.

Stride is built on the premise that shared mobility systems must be engineered with the same discipline applied to transport, communications, and energy networks. Vehicles, software, data pipelines, and enforcement mechanisms are designed as a single integrated system. This document outlines the core principles and technologies underpinning Stride’s platform, with a focus on privacy preservation, proactive safety, intelligent automation, and cross-platform accessibility.

Privacy by Architecture

Mobility data is inherently sensitive. Trip histories can reveal behaviour patterns, personal routines, work locations, health information, and social relationships. Treating this data as a by-product rather than a liability introduces systemic risk.

Stride applies privacy as an architectural constraint rather than a configurable feature. Personal identity and operational telemetry are separated at the earliest possible stage. User account data is isolated from trip data, and persistent identifiers are avoided wherever they are not strictly required for system function.

Where identifiers are necessary, they are tokenised or hashed to preserve relational structure without exposing identity. Exact GPS traces are not retained in user-linked datasets beyond operational necessity. Analytical datasets deliberately reduce temporal and spatial precision to prevent reconstruction of individual journeys.

External data sharing is deidentified by default. Personally identifiable information is stripped before export, including names, contact details, payment references, and account identifiers. Trip data is aggregated across defined zones or time windows. Where aggregation alone is insufficient, truncation or controlled noise is applied to reduce reidentification risk.

These processes are periodically reviewed against current privacy standards and emerging attack vectors, rather than static compliance checklists. The objective is not minimum compliance, but structural resistance to misuse.

Government and Council Data Access

Cities require operational visibility to manage public space. Stride supports this requirement through open, standardised data access while maintaining strict privacy boundaries.

Stride provides GBFS-compliant real-time feeds delivering vehicle availability, status, and service area definitions. Trip records and system events are posted immediately upon ride completion or state change. Historical datasets, maintenance logs, and zone configuration records are available in open formats such as JSON or CSV.

Government access rights are unrestricted for planning, regulatory, and policy use. Data may be stored, analysed, redistributed, or published by councils without licensing constraints. This access does not extend to personal surveillance. Individual rider identities are never exposed through shared datasets.

The separation between system oversight and individual tracking is explicit, documented, and enforced at the data model level.

Safety as a System Property

Safety in shared mobility cannot rely solely on static rules or rider compliance. Urban environments are dynamic, and risk fluctuates based on pedestrian density, road type, visibility, and time of day.

Stride treats safety as a system property engineered across hardware, software, and policy. Vehicles are equipped with onboard sensors and cameras feeding local processing systems capable of real-time environmental classification. This enables context-aware responses without continuous network dependency.

Speed control is dynamic. In high-risk environments such as shared paths, crossings, or pedestrian-dense zones, vehicles automatically reduce speed. These interventions occur irrespective of rider intent. The system does not wait for violations to occur.

This autoslow capability is not a fixed geofence. It responds to real-world conditions detected in real time, allowing for more granular and adaptive risk management.

Machine Learning and Risk Modelling

Stride uses machine learning to identify patterns associated with elevated risk and non-compliant behaviour. Models are trained on operational data to detect indicators such as erratic steering, repeated boundary testing, sudden acceleration in shared zones, and time-based risk profiles.

Rider interventions escalate progressively. Initial responses include in-ride warnings and enforced slow modes. Persistent high-risk behaviour results in automated restrictions or suspension. Enforcement is consistent and system-driven, reducing reliance on manual reporting or after-the-fact penalties.

Impairment detection is approached conservatively. Stride prioritises behavioural and cognitive indicators over invasive biometric methods. The goal is harm reduction, not surveillance. All such systems are validated through controlled trials and research partnerships before deployment.

Model governance is explicit. All models are versioned, monitored for drift, and retrained under defined controls. Updates are tested in shadow environments prior to activation. Safety regressions are treated as system failures requiring rollback.

Pedestrian and Accessibility Safety

Electric micromobility vehicles are inherently quiet, creating disproportionate risk for blind and low-vision pedestrians. Addressing this risk requires deliberate design, not incidental warnings.

Stride vehicles incorporate audible alert systems at low speeds, aligned in principle with Acoustic Vehicle Alerting Systems mandated for electric cars in Australia. These alerts use broadband-modulated sound profiles designed to be detectable without contributing excessive noise pollution.

Visual signalling complements audio alerts through configurable LED systems, improving detection in low-light and high-traffic environments.

Stride has collaborated with accessibility organisations, including the Canberra Blind Society, to inform design decisions and evaluate real-world effectiveness. Feedback from affected communities is treated as a primary input, not secondary validation.

Context-aware rider alerts further enhance safety. Riders receive real-time audio or visual prompts when entering high-pedestrian zones or areas requiring heightened caution. These cues are designed to prompt immediate behavioural adjustment with minimal cognitive load.

Mobile Platforms as Control Surfaces

The mobile application is a critical safety interface, not an auxiliary tool. Stride develops and maintains native iOS and Android applications with full feature parity. Platform choice should not affect access to safety features, alerts, or controls.

Both applications support real-time feedback, zone-based notifications, speed alerts, and compliance prompts. Interface design prioritises clarity under motion and reduced cognitive demand.

Native development is used for safety-critical functionality to ensure deterministic behaviour and access to platform-specific capabilities. Shared logic is employed selectively where it does not compromise performance or reliability.

Platform security standards are enforced rigorously. On iOS, this includes strict permission scoping, secure enclave usage where applicable, and substantive adherence to platform privacy requirements. On Android, this includes scoped storage, transparent foreground services, and compatibility across a diverse device ecosystem without degrading security posture.

Observability and Accountability

Automated systems operating in public space must be auditable. Stride’s backend infrastructure logs all safety-relevant decisions with contextual data. This enables review, appeal, and continuous improvement.

Enforcement actions are traceable. Model-driven interventions can be examined retrospectively to identify false positives, edge cases, or systemic bias. Black-box enforcement is explicitly avoided.

Operational analytics are provided through secure dashboards accessible to authorised officials. Metrics include fleet utilisation, compliance rates, rebalancing efficiency, maintenance cycles, and downtime intervals. These tools support proactive system management rather than reactive complaint handling.

Environmental and System Impact

Sustainability claims are quantified. Stride calculates mode shift, kilometres travelled, and emissions displacement using conservative assumptions and transparent methodologies. Environmental impact reporting is integrated into operational analytics rather than presented as standalone marketing metrics.

Shared mobility systems are evaluated on their ability to reduce car dependency, support public transport connections, and activate local economies without imposing external costs on cities.

Conclusion

Shared mobility has moved beyond experimentation. It now functions as shared urban infrastructure subject to regulatory scrutiny and public expectation.

Systems that externalise risk, monetise personal data, or rely on manual enforcement are incompatible with long-term operation. Sustainable micromobility requires restraint, transparency, and engineering discipline.

Stride’s platform is designed to reduce burden on cities, preserve individual privacy, and proactively manage risk in complex public environments. Safety is automated, privacy is structural, and accessibility is foundational.