About SpokeWatch — Adelaide bike theft intelligence & research context

This project started the day after my wife bought her first e-bike from a local marketplace. Less than 24 hours later, it was stolen from outside her workplace. The police were helpful, and the workplace security team did what they could — but the CCTV camera positioned exactly where the bike was taken was not operating effectively and, at the time, still hadn't been replaced. We also learned that outlets like Cash Converters don't take e-bikes at all because of battery risks, which narrowed the obvious resale channels and left us wondering: where do these bikes actually go?

That question — and the frustration of watching theft happen with so little follow-up intelligence — is why SpokeWatch exists. I wanted to know where thieves are stealing bikes from, and whether mapping that data could help inform riders and police about real hotspots rather than guesswork. SpokeWatch turns scattered community reports into a shared safety resource: it maps theft clusters, surfaces high-risk windows, ranks vulnerable brands, and packages prevention advice into one place — so the next rider has a better chance of keeping their bike safe, or getting it back.

SpokeWatch is not another bike registry. Bike Index, 529 Garage, BikeRegister and Australia's BikeVAULT already own recovery and have the network effects a new registry can't match.

SpokeWatch is positioned as a bicycle theft-intelligence, hotspot-mapping and prevention layer — designed to complement existing registries rather than compete with them. The focus is Adelaide first, where no locally built platform offers interactive trend and hotspot intelligence, and where SAPOL / data.sa.gov.au don't publish bicycle theft as a discrete category.

Scale and recovery

• •Victoria averages almost 12,000 reported bike thefts/year, with ~91% of cases unsolved — a figure unchanged over a decade (Bicycle Network / CSA Victoria; AIC CRG 19/84). Thefts rose ~75% over the last decade.
• •South Australia — SAPOL reported 3,180 bicycles stolen in SA in 2025, with 2,977 in 2024 and 3,009 in 2023 (ABC News, 29 Jan 2026). Thefts span the CBD, universities, pubs and private properties. Adelaide's top theft streets remain North Tce, Grenfell St, Frome Rd, Rundle Mall and Hindley St.
• •Global recovery sits around 5%, with significant underreporting (UK CSEW estimates only ~⅓ of thefts are reported).
• •United States: ~2.38 million adult bicycles stolen annually (~710 per 100,000), worth ~US$1.44 billion — bikes are ~2.5× more likely to be stolen than a car (Agarwal et al., Findings, 2025).

Behavioural and public-health impact

Theft measurably suppresses cycling. A peer-reviewed 2024 study (Cohen et al., International Journal of Sustainable Transportation, n=1,821) found 45% of victims reduced or ceased cycling after a theft, and 40% switched to less sustainable modes of transport. Recovering the stolen bike was the single biggest predictor of continued riding. This links bicycle theft directly to public-health, mode-shift and climate goals — and is why SpokeWatch frames itself as a safety and active-transport tool, not just a crime tracker.

• •Treat every report as an unverified signal. Reports are community-sourced and have not been individually checked against police records.
• •Spot patterns, not people. The map and heatmaps are designed to highlight corridors and time windows — not to identify individuals or single out properties. Locations are deliberately fuzzed.
• •Report responsibly. Only submit incidents you have direct knowledge of. Avoid naming suspects — route any identifying information to police, not to a public map.
• •Share with authorities. Local police, council safety officers and neighbourhood-watch groups are welcome to reference the public data. For aggregated, de-identified exports or research collaborations, reach out through the project's contact channels.
• •Prevention first. The analytics and toolkit sections exist to help you make smarter decisions about locks, trackers and parking — not to stoke fear.

• •Source: sample reports were manually extracted from the public Stolen Bikes South Australia Facebook page for research purposes. No private messages, identifying details or photographs have been republished here.
• •Geocoding: locations are approximated from suburb / street-level descriptions, then jittered for display. Markers do not represent exact addresses.
• •Hotspot logic: heatmaps use kernel-style density weighting on report counts. Time-of-day patterns are drawn from reported windows and should be read as indicative, not predictive.
• •Sample size: the current dataset is small and Adelaide-focused. Brand risk profiles and time/seasonality charts will become more reliable as report volume grows.

• •Australian Institute of Criminology, CRG 19/84 — bicycle theft in Victoria.
• •Cohen, Nelson, Zanotto, Fitch-Polse, Schattle, Herr & Winters (2024), "The Impact of Bicycle Theft on Ridership Behavior," International Journal of Sustainable Transportation 18(5).
• •Agarwal, Fitch-Polse, Nelson & Herr (2025), "Bicycle Theft in the US: Magnitude and Equity Impacts," Findings.
• •Milan aoristic analysis of bicycle theft (2019) — handling "stolen sometime between X and Y" time uncertainty.
• •"Cycle Thieves, We Are Watching You" RCT — signage and watching-eyes cues as theft deterrents.
• •Bicycle Network analysis of Crime Statistics Agency Victoria data (Dec 2022 and Mar 2025 updates).
• •SAPOL media releases (2019) and RAA samotor commentary — Adelaide CBD and council-level figures.

The SpokeWatch concept, identity, design and original analysis logic — including the visualisations, risk-scoring methodology and written recommendations — are the intellectual property of the project creator and are protected under applicable copyright law.

The underlying community data (individual theft reports) belongs to the original contributors and the Stolen Bikes South Australia community. SpokeWatch transforms and presents this data for safety research, but does not claim ownership of the source reports themselves.

You may reference, link to or share screenshots of public-facing outputs for non-commercial, community-safety or media-reporting purposes, provided clear attribution to "SpokeWatch" is included. Reproduction of the underlying code, data-processing methods or bulk redistribution of derived datasets for commercial purposes without prior written permission is prohibited.

Third-party trademarks and brand names referenced in the risk profiles belong to their respective owners and are used here for informational and safety purposes only.

Incident reports submitted through this prototype are displayed publicly on the map and in the analytics dashboard. Do not submit personal addresses, full names or phone numbers in report details. Map coordinates are deliberately fuzzed and aggregated for the public view; any future authority-facing exports would be de-identified and consented. No cookies or third-party trackers are used at this stage. The project is designed with the Australian Privacy Act 1988 and the 13 APPs as a guiding standard, even where not strictly applicable to a research prototype.