Smart Park Safety
Safer parks through better design, responsible technology, human partnerships, privacy protections, and public accountability.
Families should be able to enjoy parks, playgrounds, fields, trails, and community facilities with confidence. Technology may help identify emergencies, protect public property, and provide useful evidence, but cameras alone cannot maintain a park, assist a child, de-escalate a conflict, enforce every rule, provide first aid, or replace a trained person responding on site.
The goal of the Smart Park Safety plan is not to build a surveillance network. It is to create a carefully limited safety system that combines environmental design, lighting, communication, targeted technology, human response, and public oversight.
Candidate for FRRPD Board of Directors · November 3, 2026
Six Priorities for Safer Parks
Design Before Surveillance
Improve lighting, visibility, pathways, landscaping, signage, parking, emergency access, and maintenance before treating cameras as the default answer.
Targeted Technology
Use technology only where a documented safety, emergency, vandalism, or asset-protection need justifies it.
Human Response
Build clear procedures for who receives an alert, who evaluates it, who responds, and when emergency services should be contacted.
Privacy by Design
Limit collection, access, retention, sharing, analytics, and vendor control from the beginning.
Public Accountability
Publish policies, locations, purposes, retention periods, costs, access rules, audits, and program results.
Phased Investment
Begin with a limited pilot, evaluate results, and expand only if evidence, community input, funding, and public review support expansion.
Technology Should Support a Well-Designed Public Space
A camera cannot repair a broken light, clear an obstructed sightline, maintain a pathway, secure a damaged gate, clean a restroom, help a lost child, or create a positive community presence.
The first layer of park safety should be the physical environment. Before installing technology, each proposed location should be evaluated for practical improvements that may reduce risk without collecting personal information.
- Appropriate lighting
- Clear sightlines
- Maintained landscaping
- Visible entrances and exits
- Safe pathways and parking
- Legible location and emergency signage
- Reliable gates, fencing, locks, and access control where appropriate
- Timely reporting and repair of hazards
Crime-prevention and safety reviews should involve qualified staff and appropriate public-safety professionals. Community observations are valuable but are not substitutes for professional safety, engineering, accessibility, electrical, or legal evaluations.
Multiple Layers Working Together
Maintenance and Environmental Design
Lighting, landscaping, visibility, pathways, parking, signage, emergency access, facility care, and removal of conditions that make legitimate park use less comfortable.
Community Presence
Families, coaches, volunteers, recreation staff, league participants, events, programs, and active park use can create natural visibility and help residents feel connected to public spaces.
Do not suggest that residents should confront suspicious individuals or place themselves in danger.
Communication and Emergency Tools
Depending on verified needs and feasibility:
- ›Clearly marked location identifiers
- ›Emergency and non-emergency contact information
- ›Call stations or help points at selected facilities
- ›Public-address capability where justified
- ›Emergency notifications
- ›Procedures for reporting hazards, vandalism, or urgent concerns
Targeted Video Security
Potential tools may include:
- ›Fixed overview cameras
- ›Multi-sensor cameras
- ›Limited pan-tilt-zoom cameras
- ›Low-light capability
- ›Tamper alerts
- ›Object or perimeter alerts
- ›Local recording
- ›Event-based review
- ›Encrypted evidence export
Cameras should generally focus on entrances, parking areas, major circulation routes, equipment, buildings, and documented problem areas—not private property, restroom interiors, changing areas, playground close-ups, or areas where residents reasonably expect greater privacy.
Human Review and Response
Every alert system needs written procedures identifying:
- ›Who may receive an alert
- ›What training is required
- ›How an alert is verified
- ›When district staff should respond
- ›When contracted security may respond
- ›When law enforcement, fire, or emergency medical services should be contacted
- ›How false alerts are documented
- ›How evidence is preserved
- ›How incidents are reviewed afterward
Technology should not automatically dispatch law enforcement or accuse a person of wrongdoing without appropriate human verification, except where a lawful emergency system operates under separately approved procedures.
Detect Earlier. Verify Carefully. Route Responsibly.
The proposed Smart Park Safety Network would explore limited artificial-intelligence tools designed to identify specific, observable events and bring them to the attention of trained people more quickly.
The AI would not determine guilt, identify ordinary park visitors, or decide that someone “looks suspicious.” Its role would be narrower: detect a configured event, create an alert, provide useful location and visual information, and help an authorized person determine what should happen next.
When properly designed, this may help park staff and emergency partners receive clearer information sooner. However, AI can make mistakes. Every use must include testing, human judgment, documented procedures, public oversight, and the ability to disable an unreliable feature.
About this model
This is AI-assisted event detection and incident routing—not autonomous policing, predictive policing, AI law enforcement, automated criminal identification, or continuous behavioral surveillance. “Real-time” means the system can generate and transmit an alert within seconds or minutes of detecting a configured event, depending on connectivity and system performance. It does not imply a guaranteed emergency response time.
A Five-Step Workflow
Detect
A camera or connected sensor identifies a configured event that may require attention.
Package the Alert
The system creates an alert containing:
- ›Park and device location
- ›Date and time
- ›Event category
- ›Confidence level
- ›Still image or short event clip when permitted
- ›Device status
- ›A link for authorized live or recorded review
Human Verification
A trained and authorized person reviews the available information, checks for obvious errors, and determines whether the alert appears actionable.
The reviewer may be qualified FRRPD personnel, a contracted monitoring service, or another authorized partner operating under a publicly approved agreement.
Route Appropriately
After verification, the incident is directed according to an approved response matrix. Depending on the event, this may involve:
- ›FRRPD staff
- ›Maintenance personnel
- ›Contracted park-safety personnel
- ›A designated supervisor
- ›Law enforcement through the appropriate emergency or non-emergency channel
- ›Fire or emergency medical services through 911
- ›No response when the alert is determined to be false or non-actionable
Document and Close
Record:
- ›What the system detected
- ›Whether the alert was verified
- ›Who reviewed it
- ›Where it was routed
- ›What response occurred, when lawfully available
- ›Whether the detection was accurate
- ›Whether the technology or procedure needs adjustment
Different Events Require Different Responses
Conceptual only—final routing must be developed with FRRPD staff, legal counsel, emergency-communications professionals, and any participating agencies.
Potential Medical Emergency
Examples
- ›Possible person down
- ›Possible fall
- ›Manually activated emergency call station
- ›Reported serious injury
Preliminary routing
Immediate human review. If an emergency reasonably appears to exist, contact 911 for fire or emergency medical response and provide the verified location and available details.
Potential Fire or Environmental Emergency
Examples
- ›Possible smoke or flame
- ›Sudden hazardous condition
- ›Flooding or severe weather impact reported through connected systems
Preliminary routing
Immediate human review followed by 911 or the established emergency channel when an emergency reasonably appears to exist.
Potential Violence or Immediate Threat
Examples
- ›Reported assault
- ›Possible active violence
- ›Manually activated emergency alert
- ›Other observable conduct suggesting an immediate danger
Preliminary routing
Priority human review and contact with 911 when immediate police, fire, or medical assistance may be required.
Do not claim that AI can reliably identify a weapon or determine criminal intent.
Possible Vandalism or Property Damage
Examples
- ›Possible damage to a building, light, restroom, playground, field, gate, or other public asset
- ›Activity within an approved restricted area after hours
Preliminary routing
Human review followed by FRRPD staff, contracted safety personnel, or the appropriate non-emergency law-enforcement channel based on the apparent seriousness and adopted procedures.
Maintenance or Facility Concern
Examples
- ›Lighting outage
- ›Gate or access failure
- ›Camera tampering
- ›Water accumulation
- ›Equipment or system failure detected by a connected sensor
Preliminary routing
Create a maintenance notification or work request for authorized district review. Do not route ordinary maintenance alerts to emergency services.
Crowding or Unusual Activity
Examples
- ›Attendance exceeding a configured operational threshold
- ›Unexpected activity in a closed facility
- ›Large movement near a controlled entrance
Preliminary routing
Human review only. Crowding or presence must not automatically be treated as wrongdoing.
False or Non-Actionable Alert
Examples
- ›Wildlife
- ›Weather
- ›Shadows
- ›Normal recreation
- ›Authorized staff
- ›Maintenance activity
- ›Camera movement
- ›System error
Preliminary routing
No outside response. Record the result and use it to improve settings, training, or system evaluation.
AI Would Not Replace 911 or Emergency Dispatch
The proposed system would not operate as an independent emergency-dispatch center. Police, fire, and medical responses must continue through the appropriate emergency-communications process.
AI may help an authorized reviewer recognize an event and provide location or visual context sooner. Trained 911 professionals and responding agencies—not an FRRPD camera system—determine the appropriate emergency response.
Automatic transmission directly into a public-safety answering point should not be enabled unless the receiving agency agrees, the integration has been legally and technically approved, responsibilities are documented, testing is completed, and the public process has occurred.
What the AI May—and May Not—Evaluate
Permitted Pilot Functions
- Line crossing at a closed or controlled area
- Entry into a defined restricted zone
- Possible person-down event
- Possible smoke or flame indicator
- Camera tampering
- Unusual crowd density based on a neutral numerical threshold
- Object or vehicle presence in a restricted operational area
- Lighting, power, network, gate, or connected-sensor failure
- Motion around specified public assets during closed hours
Prohibited Initial Functions
- Facial recognition
- Automated license-plate recognition
- Emotion recognition
- Race, ethnicity, age, disability, religion, gender, or political-activity classification
- Predictive policing
- Generalized behavioral profiling
- Automated identification of “suspicious people”
- Social-media matching
- Tracking ordinary visitors between parks
- Inferring criminal intent
- Fully automated police dispatch
- Creating watchlists of ordinary residents
- Assigning a criminal-risk score to a person
A Person Must Remain Accountable
- 1.A named human role responsible for every alert queue.
- 2.Written authority defining who may review and route alerts.
- 3.Training before system access.
- 4.Human verification before external escalation, except a direct user-activated emergency device operating under an approved procedure.
- 5.An access and action log.
- 6.Supervisor review of significant alerts.
- 7.A process for reporting incorrect alerts.
- 8.A process for correcting configuration problems.
- 9.The ability to immediately disable an analytics feature.
- 10.No disciplinary, enforcement, or accusatory decision based solely on AI output.
- 11.No assumption that an alert proves wrongdoing.
The System Must Prove It Is Useful
Before activation
- ›Test every analytics function at the proposed location.
- ›Test daytime, nighttime, weather, seasonal, and high-activity conditions.
- ›Establish preliminary confidence thresholds.
- ›Test common causes of false alerts.
- ›Verify privacy masking.
- ›Verify time, location, and device information.
- ›Test the human-review and routing process.
- ›Conduct planned tests with participating responders where authorized.
- ›Document known limitations.
During operation, measure
- ›Alerts by category
- ›Verified alerts
- ›False-positive rate
- ›Known false negatives when discoverable
- ›Average time from detection to human review
- ›Average time from verification to routing
- ›System uptime
- ›Network outages
- ›Events missed because of coverage limitations
- ›Alerts dismissed by reviewers
- ›Alerts routed to each responder category
- ›Emergency contacts resulting from alerts
- ›Maintenance requests generated
- ›Complaints
- ›Privacy incidents
- ›Cybersecurity events
- ›Cost per verified event
Do not use a single accuracy percentage without explaining the event type, location, testing conditions, and sample size.
An analytics feature that produces excessive false alerts, creates inequitable effects, fails cybersecurity review, or does not provide measurable operational value should be adjusted, suspended, or removed.
The Safety Plan Must Work When Technology Does Not
Plan for
- •Power failure
- •Internet or cellular outage
- •Camera obstruction
- •Vandalism
- •Server or storage failure
- •Software outage
- •Cyberattack
- •Expired certificate or license
- •Monitoring-service interruption
- •Staff vacancy
- •Failure to acknowledge an alert
Require
- ›Device-health monitoring
- ›Escalation for unacknowledged critical alerts
- ›Documented maintenance contacts
- ›Backup communications procedures
- ›Local recording where appropriate
- ›Recovery priorities
- ›Regular system tests
- ›Public reporting of material outages when disclosure does not create a security risk
The absence of an alert does not prove that a park is safe or that no incident occurred.
Useful Tool—Not a Complete Safety Department
What Responsible Technology May Help With
- Documenting vandalism or property damage
- Reviewing a reported incident
- Alerting designated personnel to activity in a defined area
- Monitoring entrances, buildings, parking areas, or equipment
- Identifying lighting, visibility, or maintenance concerns
- Preserving evidence for an authorized investigation
- Supporting emergency communication
- Helping evaluate when and where additional human presence may be needed
What Technology Cannot Replace
- First aid
- De-escalation
- Visitor assistance
- Welfare checks
- Rule education
- On-site judgment
- Maintenance
- Youth supervision
- Emergency medical response
- Fire response
- Law-enforcement authority
- A trusted human presence
A camera may record an emergency without preventing it. A responsible safety plan must explain who responds.
Public Safety Requires Public Trust
- 1.
Defined Purpose
Every device must have a documented purpose connected to a verified park-safety or asset-protection need.
- 2.
Public Location List
Publish the general location, purpose, device type, and responsible authority for each approved installation, except where limited information is legally necessary to protect system security.
- 3.
Visible Notice
Post understandable signs informing visitors that limited video security is in use and directing them to the complete policy.
- 4.
Data Minimization
Collect only the information reasonably needed for the approved purpose.
- 5.
Short Default Retention
Establish a short default retention period. Use 30 days or less as the preliminary policy goal for ordinary video unless legal counsel, operational evidence, or a specific investigative need supports a different period. Flagged evidence may be preserved longer under documented evidence-retention procedures.
- 6.
Limited Access
Only trained and specifically authorized personnel should access recordings.
- 7.
Access Logs
Record who accessed footage, when it was accessed, and the documented reason.
- 8.
Encryption and Cybersecurity
Require encryption, strong authentication, secure updates, network separation where appropriate, backup controls, vulnerability management, and a documented breach-response plan.
- 9.
Controlled Sharing
Do not provide unrestricted access to outside agencies or vendors. Sharing should require a documented, lawful purpose and comply with applicable California law.
- 10.
No Sale or Commercial Use
Prohibit the sale, advertising use, commercial exploitation, or unrelated analysis of park-security data.
- 11.
Independent Auditing
Conduct recurring audits of access, retention, sharing, cybersecurity, vendor compliance, and adherence to the adopted policy.
- 12.
Public Reporting
Publish aggregate information about requests, approved disclosures, denied requests, retention exceptions, security incidents, complaints, and program performance without exposing private or investigative details.
Technology Must Not Decide Who “Looks Suspicious”
- Do not include facial recognition in the initial program.
- Do not identify people based on biometric characteristics.
- Do not use analytics to classify people by race, ethnicity, religion, disability, age, clothing, political activity, or other protected or sensitive characteristics.
- Do not create generalized movement profiles of ordinary park visitors.
- Do not use predictive labels claiming that a person is likely to commit misconduct.
- Do not automatically treat presence after an alert as evidence of wrongdoing.
- Require separate public review, legal analysis, civil-rights evaluation, and Board approval before considering any future biometric capability.
Do Not Quietly Turn Park Cameras into a Vehicle-Tracking Network
Automated license-plate readers are different from ordinary security cameras. They convert license plates into searchable records connected to a time and location. This can create a detailed history of vehicle movements and raises significant legal, privacy, cybersecurity, vendor-control, accuracy, and data-sharing concerns.
Recent California enforcement actions involving the sharing of license-plate data demonstrate why public agencies must not assume that a vendor's default network, sharing settings, or retention rules comply with local expectations or California law.
- ✕Do not include ALPR technology in the initial Smart Park Safety pilot.
- ✕Do not connect ordinary park cameras to a license-plate recognition network.
- ✕Do not enable nationwide or unrestricted data sharing.
- ✕Do not permit a vendor to change sharing settings without written authorization.
- ✕Do not share data with federal or out-of-state agencies unless a verified legal requirement permits or compels the specific disclosure.
- ✕Do not allow a vendor to sell, repurpose, aggregate, or commercially exploit data.
- ✕Require FRRPD to retain practical control of its data, settings, permissions, exports, and deletion.
- ✕Require written policies, training, auditing, public comment, cybersecurity review, and legal review before any later ALPR proposal.
- ✕Require any future ALPR proposal to return to the Board as a separate public decision.
No vendor endorsement.
This policy does not endorse Flock Safety or any competing vendor. Products from Flock, Axis Communications, Hanwha Vision, Uniview, Milestone Systems, Genetec, and other companies may be examined during future competitive research. Mentioning a company does not recommend procurement, and no vendor should be selected without technical evaluation, privacy review, competitive purchasing, legal compliance, lifecycle-cost analysis, and public accountability.
The District Should Control Its Safety System
- ›Open, standards-based cameras when feasible
- ›A competitively selected video-management platform
- ›Local or district-controlled recording
- ›Encrypted event-based backup when justified
- ›Exportable evidence in standard formats
- ›Documented administrative ownership
- ›No undisclosed third-party data sharing
- ›Clear end-of-contract deletion and data-return requirements
- ›Ability to replace components without replacing the entire network
- ›Cybersecurity updates and support throughout the expected service life
Solar or wireless equipment may be considered where trenching power or data lines is impractical, but reliability, maintenance, connectivity, vandalism exposure, battery replacement, and lifecycle cost must be studied before selection.
Smart Park Network vs. Full-Time Ranger Division
Preliminary figures — not bids or budgets.
The following figures are preliminary 2026 planning ranges—not vendor bids, FRRPD budget figures, employment offers, approved spending, or final cost estimates. Actual costs require a districtwide needs assessment, coverage plan, labor classification, legal review, insurance analysis, site engineering, competitive proposals, and staff evaluation.
Technology estimate includes
Equipment, installation, power and connectivity, recording, software, cybersecurity, training, signage, monitoring or alert review, maintenance, and replacement reserves.
Ranger estimate includes
Wages, payroll expenses, pensions and benefits, supervision, vehicles, fuel, radios, uniforms, training, background investigations, insurance, workers' compensation, scheduling, overtime, administration, facilities, and equipment.
Smart Park estimates must also include AI-related costs
AI analytics licenses, edge-processing devices or capable cameras, video-management integration, alert-management software, monitoring or human-review labor, network and cellular costs, cybersecurity, software support, testing and calibration, false-alert review, training, maintenance, replacement, data storage, and privacy and legal review.
Do not treat AI software as a one-time camera expense. Recurring analytics, monitoring, connectivity, cybersecurity, maintenance, and system-administration costs must be included in the total cost of ownership.
AI-assisted analytics and human alert review are included within the displayed preliminary ranges as preliminary allowances. The ranges were not changed by this clarification.
Current California public-agency examples show ranger salaries ranging approximately from the mid-$40,000s to the mid-$90,000s annually, with supervisory compensation potentially exceeding $125,000 before benefits and operating costs.
Option 1 — Limited Smart Park Pilot
Illustrative Scope
- ›Approximately 4–6 priority locations
- ›Approximately 12–18 targeted cameras or sensors
- ›Limited local recording
- ›Signage and privacy policy
- ›Event-based alert review
- ›Lighting, communication, or access improvements where included in the approved pilot
- Preliminary startup:
- $175,000–$400,000
- Annual operating:
- $75,000–$175,000
- 5-year total:
- $550,000–$1,275,000
Coverage: Technology coverage at selected locations. Does not provide a continuous human presence.
Option 2 — Broader Smart Park Safety Network
Illustrative Scope
- ›Approximately 8–12 priority facilities or park areas
- ›Approximately 24–40 targeted camera views
- ›Local recording and video management
- ›Limited analytics
- ›Cybersecurity controls
- ›Signage
- ›Training
- ›Maintenance and replacement reserve
- ›Event review and defined response procedures
- Preliminary startup:
- $400,000–$850,000
- Annual operating:
- $150,000–$350,000
- 5-year total:
- $1,150,000–$2,600,000
Coverage: Broader technology coverage, but still requires people to review, respond, maintain equipment, and assist visitors.
Option 3 — Starter Full-Time Ranger Division
Illustrative Scope
- ›One supervisor or lead
- ›Four full-time rangers
- ›Limited seven-day and evening coverage
- ›Vehicles, radios, uniforms, equipment, training, administration, and relief coverage
- Preliminary startup:
- $250,000–$550,000
- Annual operating:
- $750,000–$1,250,000
- 5-year total:
- $4,000,000–$6,800,000
Coverage: Provides a human presence during scheduled shifts but does not create continuous coverage at every park or guarantee 24-hour response.
Option 4 — Expanded Ranger Division
Illustrative Scope
- ›Supervision
- ›Approximately 8–10 ranger positions
- ›Relief coverage
- ›Broader evening, weekend, and emergency scheduling capacity
- ›Multiple vehicles and supporting equipment
- Preliminary startup:
- $450,000–$900,000
- Annual operating:
- $1,500,000–$2,500,000
- 5-year total:
- $7,950,000–$13,400,000
Coverage: Provides substantially more human coverage but still cannot place a ranger at every facility at all times.
A single position requiring continuous 24-hour staffing takes more than four full-time-equivalent employees before accounting for vacations, training, illness, injuries, vacancies, court appearances, administrative work, or two-person response requirements. A genuine 24-hour ranger division therefore requires substantially more staffing than simply hiring one or two rangers.
Do not label the Smart Park options as equivalent to a ranger division. They provide different services.
Each Model Solves a Different Part of the Problem
Technology-Centered Model
Advantages
- Wider fixed-location coverage
- Lower preliminary operating cost than a ranger division
- Evidence collection
- Scalable pilot
- Can operate outside staffed hours
Limitations
- No physical intervention
- No first aid or visitor assistance
- False alerts
- Cybersecurity and privacy risks
- Ongoing monitoring and maintenance costs
- Coverage limited to installed locations
Ranger-Centered Model
Advantages
- Visible human presence
- Visitor assistance
- Education and de-escalation
- On-site judgment
- First aid if appropriately trained
- Direct observation beyond camera views
Limitations
- Higher recurring cost
- Hiring and retention challenges
- Vehicles, training, pensions, insurance, and supervision
- Limited coverage during each shift
- Legal authority and classification must be established
- One or two rangers cannot cover every facility continuously
Hybrid Model
Advantages
- Uses technology for broader awareness
- Preserves human judgment and response
- Allows phased staffing
- Can prioritize high-use times and locations
- May provide better coverage per public dollar
Limitations
- More complex coordination
- Still requires operating funding
- Requires clear roles and procedures
- Technology and staffing must both be evaluated
- Poorly designed programs can inherit the weaknesses of both systems
Begin with a Privacy-First Hybrid Pilot
The recommended direction is not an immediate districtwide surveillance rollout or the immediate creation of a large ranger division.
The responsible starting point is a limited hybrid pilot built around documented needs, improved lighting and environmental design, targeted technology, strict privacy controls, and clearly identified human-response partners.
The pilot should:
- 1.Select a limited number of locations through documented criteria and public discussion.
- 2.Address maintenance, lighting, visibility, signage, and access concerns before or alongside technology.
- 3.Exclude facial recognition and ALPR.
- 4.Use short retention and district-controlled access.
- 5.Establish written response procedures.
- 6.Provide public notice before activation.
- 7.Publish actual startup and operating costs.
- 8.Track false alerts, incidents, response times, downtime, privacy complaints, access logs, and public feedback.
- 9.Operate for a defined evaluation period.
- 10.Return to the Board for public review before expansion.
Possible human-response components to study include:
- ›Existing FRRPD staff roles
- ›Contracted security during identified high-risk periods
- ›Park hosts or ambassadors
- ›Seasonal or part-time safety personnel
- ›Appropriate coordination with the Oroville Police Department
- ›Appropriate coordination with the Butte County Sheriff's Office
- ›Community-service and volunteer programs that do not exercise enforcement authority
- ›A future ranger or park-safety classification if evidence and funding justify it
Do not claim that any outside agency has committed to participate.
Where a Pilot Might Be Considered
No pilot parks are predetermined. Locations should be selected through the criteria below, not announced in advance.
- ›Documented incident patterns
- ›Vandalism or repeated property damage
- ›High visitor use
- ›Evening activity
- ›Limited visibility
- ›Emergency-access concerns
- ›Existing lighting and power
- ›Cost of installation
- ›Network availability
- ›Privacy impact
- ›Proximity to private property
- ›Community feedback
- ›Staff workload
- ›Ability to measure results
Residents have previously raised safety concerns about locations including Martin Luther King Jr. Park and Bedrock Park. Those observations may justify further review, but they do not replace verified incident data, staff analysis, or public site evaluation.
A Public Path from Proposal to Program
Phase 1 — Listen and Document
- ›Gather resident and staff concerns.
- ›Review available incident and maintenance information.
- ›Identify safety-design improvements.
- ›Consult privacy, accessibility, cybersecurity, legal, and public-safety expertise.
Phase 2 — Design the Pilot
- ›Establish locations and purposes.
- ›Complete site and network assessments.
- ›Draft privacy and usage policies.
- ›Define response procedures.
- ›Publish preliminary lifecycle costs.
Phase 3 — Public Review and Procurement
- ›Present the pilot publicly.
- ›Receive resident feedback.
- ›Complete legal review.
- ›Use competitive procurement.
- ›Evaluate vendors on security, privacy, interoperability, reliability, support, and total cost—not purchase price alone.
Phase 4 — Limited Deployment
- ›Install approved equipment.
- ›Test coverage and privacy masking.
- ›Train authorized users.
- ›Post public notice.
- ›Test alert and emergency procedures before operation.
Phase 5 — Evaluate Before Expanding
- ›Measure performance.
- ›Audit access and retention.
- ›Publish costs.
- ›Review complaints and false alerts.
- ›Compare outcomes with other safety investments.
- ›Continue, modify, expand, or discontinue only through a public decision.
How Results Should Be Tracked
- ›Documented incidents at pilot locations
- ›Calls for service where data are lawfully available
- ›Vandalism and repair costs
- ›Equipment uptime
- ›Alert volume
- ›False-alert rate
- ›Verified-event rate
- ›Response process and timing
- ›Unauthorized-access attempts
- ›Data-sharing requests
- ›Retention exceptions
- ›Privacy complaints
- ›Cybersecurity incidents
- ›Visitor perceptions of safety
- ›Evening and program use
- ›Actual startup and annual cost
- ›Cost per verified safety event
- ›Staff workload
- ›Evidence requests that produced useful information
A reduction or increase in reported incidents must be interpreted carefully. Increased reporting may reflect improved awareness or public confidence rather than worsening conditions.
One Director Cannot Create a Safety Department Alone
If elected, Dacoda Quinn would serve as one member of the FRRPD Board of Directors. He could request information, ask questions, propose public discussion, recommend safeguards, review agreements, evaluate budgets, and vote on matters properly before the Board.
Purchasing technology, hiring personnel, establishing a ranger classification, entering agreements, directing staff, sharing information, or creating enforcement authority requires the appropriate public process.
Any ranger proposal would require analysis of the position's legal authority, duties, labor classification, training, background requirements, supervision, insurance, workers' compensation, equipment, pension obligations, policies, and relationship with law-enforcement agencies.
A title such as "park ranger" does not automatically grant peace-officer or law-enforcement authority.
How These Preliminary Ranges Were Developed
The ranger ranges use current California public-agency salary examples, supervisory compensation, and estimated employer costs for payroll, benefits, retirement, vehicles, equipment, training, insurance, and administration.
The technology ranges are preliminary planning allowances for equipment, video management, installation, trenching or wireless connections, power, recording, storage, cybersecurity, monitoring, maintenance, replacement, signage, training, and contingency.
Actual costs could fall below or above these ranges depending on:
- ›Number of parks
- ›Number and type of devices
- ›Power availability
- ›Trenching
- ›Solar requirements
- ›Wireless or fiber connectivity
- ›Video retention
- ›Monitoring model
- ›Installation conditions
- ›Cybersecurity requirements
- ›Staffing classifications
- ›Benefit obligations
- ›Vehicle requirements
- ›Training
- ›Insurance
- ›Procurement results
Source links
- California Department of Justice — ALPR Guidance and Privacy Requirements
- California Department of Justice — 2025 El Cajon ALPR Enforcement Action
- CalPERS — 2026–27 Employer Contribution Rates
- Riverside County — 2026 Park Ranger Salary Example
- City of Los Angeles — 2026 Park Ranger Salary Example
- City of Sacramento — 2026 Chief Park Ranger Salary Example
- Axis Communications — Security-System Total Cost of Ownership
- National Institute of Standards and Technology — AI Risk Management Framework
A voluntary framework for governing, mapping, measuring, and managing risks associated with artificial-intelligence systems.
- National 911 Program — Calling 911
Federal public guidance explaining that 911 call-takers gather the location, nature, and details of an emergency to help direct the appropriate police, fire, or medical response.
Sources are provided for methodology and public education. They do not constitute vendor endorsements or proof of FRRPD's eventual costs. Update the model when verified FRRPD information and competitive proposals become available.
What Would Help You Feel Safer in Our Parks?
Park safety affects families, athletes, coaches, seniors, neighbors, staff, and every resident who wants to enjoy public space. This plan should be shaped through honest discussion about safety needs, privacy, costs, staffing, technology, and the kind of public presence residents want.
- ›Which locations or conditions concern you?
- ›Would better lighting or visibility help?
- ›Where would a human presence matter most?
- ›What privacy limits should be required?
- ›Would you support a limited pilot?
- ›What information should be included in public reports?
Safer parks require more than cameras and more than slogans. They require sound design, responsible people, limited and accountable technology, public trust, and an honest plan for who responds when help is needed.
