About this skill

Problem

In industrial parks, communities, campuses, and restricted roads, e-bikes and electric scooters are often parked illegally or driven into restricted areas. Manual review of surveillance footage is slow, and it is hard to keep consistent violation counts and auditable reports. This skill turns that review into a repeatable detection workflow.

How it works

Input: accepts surveillance video or still images, commonly mp4, avi, mov, jpg, png, and jpeg, with files usually capped around 10MB.
Detection targets: identifies e-bikes and electric scooters in the frame and evaluates whether they appear in restricted zones.
Area types: distinguishes parking-lot, community, campus, road, and other for scenario-specific reporting.
Output: produces a structured report with monitoring area details, detection statistics, vehicle counts, violation level, and recommended actions.
History lookup: historical reports should be fetched from the cloud interface and rendered as a Markdown table, rather than assembled from local notes.

Boundaries

The skill is best used as a security-management aid, not a replacement for manual verification. Privacy, evidence handling, and enforcement decisions still require human review and compliant processes. Keep the analysis inside the skill's built-in scripts instead of generating ad-hoc analysis code.

Screenshots

Use Cases

A park security team uses a video clip to identify e-bikes in a restricted zone and produce violation counts.
A community duty officer checks an image to see if an electric scooter is in the parking area and get recommended actions.
A site safety manager retrieves historical detection reports and lists report names with violation counts.
A campus operations team detects a questionable road clip for the campus scenario and obtains a violation level.

Best For

Park security lead: turns a restricted-zone video into vehicle counts and a violation-level report.
Community duty officer: checks a still image to confirm whether an electric scooter is in the parking area.
Site safety manager: retrieves historical reports and lists report names with violation counts in a table.
Campus operations on-duty staff: detects a questionable road clip under the campus scenario and gets a violation level.