NU FAIRVIEW
School of Engineering and Technology
Bachelor of Science in Information Technology
Specialization in Mobile and Internet Technology

SAFEWATCH: AN AI-ENABLED CCTV MONITORING, INCIDENT RESPONSE, AND PUBLIC ANNOUNCEMENT SYSTEM FOR BARANGAY HOLY SPIRIT

“Good ideas get to work.”

Midterm Documentation Submitted in Partial Fulfillment
of the Requirements for Technopreneurship (TENTREP)

By:

Andam, John Carlo D.

Baconawa, Cyrus Jeshurun Jacob M.

Dela Pena, Merl Angelo S.

Ilio, Benedict L.

Sanchez, Denver D.

Adviser: John Rey P. Radoc
Subject Teacher: Herminiño C. Lagunzad

Academic Year 2026-2027

TABLE OF CONTENTS

Chapter 1. Startup Branding 1

Chapter 1. Startup Branding

1.1 Company Name

The official name of the startup is Obsidian IT Solutions. The word “Obsidian” refers to a naturally formed volcanic glass recognized for its strength, sharpness and distinctive dark color. These qualities reflect the company’s commitment to developing precise, resilient, and dependable technology solutions. The term “IT Solutions” communicates the company’s broader purpose of designing systems that respond to practical organizational and community needs.

Obsidian IT Solutions is the company behind SafeWatch, its flagship project. SafeWatch is an AI-enabled, audio-assisted CCTV monitoring and incident response system designed initially for Barangay Holy Spirit. This distinction establishes Obsidian IT Solutions as the startup and SafeWatch as a technology solution developed under its brand. The company name is sufficiently broad to accommodate the future development of web, mobile, artificial intelligence, data management, and community-oriented technology solutions beyond the SafeWatch project.

1.2 Company Logo

The Obsidian IT Solutions logo serves as the primary visual identity of the company. Its design combines a faceted obsidian shield, a gold camera-aperture symbol, a bold geometric wordmark, and the tagline “Good ideas get to work.” Together, these elements represent technological strength, intelligent observation, responsible protection, and the company’s commitment to creating solutions with practical value.

Figure 1. SafeWatch Official Logo Design

Meaning of the Logo

The logo combines a faceted obsidian shield with a gold camera-aperture symbol to represent the identity and purpose of Obsidian IT Solutions. The shield symbolizes protection, resilience, operational security, and the company’s responsibility to safeguard the people, information, and organizations supported by its systems, while its angular facets reflect the strength and distinctive appearance of an obsidian. At its center, the camera aperture represents SafeWatch, the company’s flagship project, as well as CCTV monitoring, computer vision, situational awareness, and AI-assisted incident detection. Together, these elements communicate that Obsidian IT Solutions uses intelligent technology to support protection, responsible decision-making, and authorized human control while maintaining a corporate identity suitable for future IT solutions.

Color Psychology

Obsidian Black and Gray: Obsidian Black and Gray: Reflect the natural dark appearance of obsidian and give the brand a bold, modern, and distinctive visual identity.

Gold: Creates a strong contrast with the darker colors and draws attention to the camera-aperture symbol, emphasizing it as the logo’s focal point.

White: Improves readability and provides a clean visual balance against the darker elements.

Symbolism

Faceted Obsidian Shield: Represents protection, resilience, operational security, and the technical strength of Obsidian IT Solutions.

Camera Aperture: Represents SafeWatch, CCTV monitoring, vision, AI-assisted detection, and situational awareness.

Geometric Facets: Represents the different technologies, processes, and professional disciplines working together to create one dependable solution.

Bold Wordmark and Balanced Composition: Communicate confidence, reliability, and modern engineering.

1.3 Tagline

“Good ideas get to work.”

The tagline expresses Obsidian IT Solutions’ commitment to turning useful ideas into practical technology. “Good ideas” refers to recognizing real needs and designing thoughtful ways to address them. “Get to work” emphasizes developing solutions that people can use to carry out their responsibilities.

SafeWatch reflects this approach by helping barangay personnel detect incidents and coordinate responses. The tagline also remains broad enough to represent future projects developed by Obsidian IT Solutions.

1.4 Vision

To become a trusted IT solutions company known for building smart, secure, and reliable technologies that help organizations work better and serve their communities with confidence.

1.5 Mission

Obsidian IT Solutions designs and develops practical information-technology solutions that combine artificial intelligence, web and mobile technologies, secure data management, and human-centered workflows. Through SafeWatch and future projects, the company seeks to improve situational awareness, operational coordination, accountability, and community service while protecting data, preserving authorized human control, and building on the systems and resources already available to its clients.

1.6 Core Values

Vision: Identifying real-world needs and creating forward-looking technology solutions that are purposeful, adaptable, and capable of producing long-term benefits.

Vigilance: Maintaining awareness, preparedness, responsibility, and careful attention to system performance, operational risks, data protection, and situations requiring authorized human review.

Value: Delivering reliable, accountable, and sustainable solutions that improve operations, support users, make responsible use of existing resources, and contribute positively to the community.

Chapter 2. Project Overview

SafeWatch is an AI-enabled CCTV monitoring and incident-response system for Barangay Holy Spirit. It will analyze CCTV feeds, generate AI detections and recommended threat levels, support incident verification, facilitate the existing manual dispatch workflow, provide time-based automatic dispatch for unattended detections, monitor authorized responder locations, collect incident statements and formal reports, and prepare localized emergency announcements for confirmed incidents.

SafeWatch will combine decision support with a limited, rule-based automatic-dispatch function. Artificial intelligence will not independently verify incidents or activate public speakers. When an alert is noticed within the configured period, the Command Center Personnel may verify and forward the incident to the appropriate Purok Leader, the Purok Leader may dispatch Barangay Tanods, and the Command Center Personnel may dispatch the Barangay Task Force when necessary. When an alert is not noticed within the configured period, SafeWatch will automatically dispatch the nearest and appropriate Barangay Tanods or Task Force for Level 1, Level 2, and Level 3 detections. Public announcements will remain separate from responder dispatch and will require Command Center verification and authorization.

2.1 Problem

Barangay Holy Spirit faces challenges in monitoring its CCTV network of approximately 200 cameras. Continuous manual observation makes it difficult to maintain equal attention across all feeds, especially during simultaneous incidents, shift transitions, and busy periods. This may delay the recognition of physical altercations, prolonged loitering, curfew-related situations, fire, or smoke.

Incident details, responder coordination, location tracking, status updates, and reports may also be handled through separate processes. This reduces efficiency and makes it harder to maintain complete and consistent incident records. An AI detection may also remain unnoticed when Command Center personnel are handling several monitoring and coordination duties at the same time. A structured system is needed to notify the Command Center, provide a configurable period for personnel to act, initiate the appropriate fallback dispatch when an alert remains unattended, and support tracking and reporting.

The barangay also requires controlled public emergency announcements. Localized announcements must remain subject to human verification and authorization to prevent unverified AI detections from triggering unnecessary broadcasts. SafeWatch must support these functions without replacing the barangay's existing CCTV and public-address systems.

2.2 Solution

SafeWatch enhances the barangay's existing CCTV infrastructure through an AI-enabled application layer, rather than replacing current systems. CCTV camera feeds are analyzed for configured incident categories, and detected events generate alerts with a recommended threat level: Level 1 Advisory, Level 2 Priority, or Level 3 Critical.

The Command Center Personnel PC application supports alert review, incident verification, threat classification, forwarding verified incidents to the appropriate Purok Leader, manual dispatch actions, and monitoring of automatic-dispatch status. When an alert is noticed within the configured period, Purok Leaders may dispatch recommended Barangay Tanods and the Command Center may separately dispatch Barangay Task Force members when needed. When an alert remains unnoticed, SafeWatch automatically dispatches the nearest and appropriate Barangay Tanods or Task Force for Level 1, Level 2, and Level 3 detections

The mobile applications support responder GPS tracking, dispatch information, status updates, incident statements, and formal reporting based on user role. The IT Department Administrator PC application manages accounts, roles, cameras, locations, curfew schedules, detection categories, speaker units, settings, and connectivity.

For confirmed fire incidents, SafeWatch prepares a localized prerecorded announcement for Command Center authorization. A centralized backend and database manage detections, incidents, dispatches, GPS records, reports, and announcement activity.

Figure 2. SafeWatch End-to-End System Architecture

2.3 Target Users

Primary Users – Command Center Personnel: Review CCTV feeds and AI detections, verify incidents, finalize threat levels, forward verified incidents to Purok Leaders, perform authorized manual dispatch actions, monitor manual and automatic responses, and authorize localized fire announcements.

Primary Users – IT Department Administrators: Manage authorized accounts, roles, cameras, locations, curfew schedules, detection categories, speaker units, approved recordings, system settings, connectivity status, and the configurable automatic-dispatch period.

Primary Users – Purok Leaders: Share authorized on-duty GPS location, review verified incidents forwarded by the Command Center, dispatch recommended Barangay Tanods, review Tanod incident statements, and prepare formal incident reports.

Primary Users – Barangay Tanods: Share authorized GPS location, receive manual or automatic dispatch assignments, view incident details and maps, update response status, and submit incident statements with supporting photos.

Primary Users – Barangay Task Force Members: Receive manual or automatic Command Center dispatch assignments, view incident information, share authorized GPS and response status, and submit formal incident reports.

Chapter 3. Innovation

3.1 Innovation Statement

SafeWatch introduces an AI-enabled CCTV monitoring and incident-response system that supports Barangay Holy Spirit’s existing public-safety operations. It combines CCTV camera feed AI detection, three-level threat recommendations, human verification, manual and time-based automatic dispatch, GPS monitoring, structured reporting, and localized fire announcements into one controlled workflow.

Unlike systems focused only on CCTV viewing or automatic detection, SafeWatch is designed around the barangay’s actual response structure. The Command Center Personnel may verify and forward incidents, Purok Leaders may manually dispatch Barangay Tanods, and the Command Center may manually dispatch Barangay Task Force members. If an AI detection is not noticed within the configured period, the system applies the approved threat-level mapping and automatically dispatches Barangay Tanods for Levels 1 and 2 or Barangay Task Force members for Level 3.

Its main innovation is the combination of AI decision support with a limited, rule-based fallback dispatch. Incident verification, final threat classification, manual dispatch actions, and public announcements remain associated with authorized roles, while unattended detections follow the configured period and approved automatic-dispatch mapping. Public announcements remain separate and always require Command Center verification and authorization.

3.2 Design Thinking Summary

Empathize

The team reviewed Barangay Holy Spirit’s current operating environment, including its CCTV network, responder coordination, and public-address capabilities. The review considered how Command Center Personnel, IT Department Administrators, Purok Leaders, Barangay Tanods, and Barangay Task Force members support monitoring and incident response.

Define

Findings: Approximately 200 CCTV cameras require continuous monitoring, making it difficult to maintain equal attention across all feeds. Incident details, responder updates, and reports may also be handled through separate processes, reducing efficiency and traceability.

Problem Statement: Barangay Holy Spirit requires a controlled decision-support system that can identify possible incidents, provide a threat-based dispatch for unattended alerts, maintain incident records, and assist with localized public announcement under Command Center authorization.

Ideate

The team considered approaches such as increasing monitoring personnel, using stand-alone detection systems, deploying a GPS-only responder application, or replacing existing platforms with a centralized system. The selected approach was a modular application layer combining AI-assisted detection, human verification, a configurable response period, manual and time-based automatic dispatch, GPS monitoring, structured reporting, and controlled audio announcements while preserving the barangay’s existing infrastructure.

Project

The planned prototype consists of Command Center Personnel and IT Department Administrator PC applications; mobile applications for Purok Leaders, Barangay Tanods, and Barangay Task Force members; an AI detection module; configurable waiting-period and automatic-dispatch processing; GPS and mapping services; notifications; reporting functions; and a centralized backend and database.

Figure 3. SafeWatch Functional Prototype

Test - Planned Validation

Controlled testing will evaluate authentication, role permissions, AI detection, incident verification, threat-level modification, the configurable response period, manual and automatic dispatch, GPS updates, notifications, status changes, reporting, and the separation of responder dispatch from public announcements. Simulated noticed and unattended incidents will also be used to identify false positives, missed detections, timing errors, permission errors, camera compatibility issues, and performance limitations before prototype acceptance.

3.3 Proposed Innovative Features

Table 1. Proposed Innovative Features Matrix

3.4 Innovation Value

Customer Value: Provides Barangay personnel with a structured workflow for reviewing AI detections, applying a defined fallback dispatch when an alert remains unattended, coordinating responders, and maintaining complete records. Barangay Tanods and Barangay Task Force members receive the applicable dispatch details, maps, status controls, and reporting requirements through the mobile application.

Social Value: Supports earlier awareness of public-safety incidents and a defined response to unattended alerts while preserving human judgment for verification, manual actions, and public-announcement authorization.

Operational Value: Connects detections, the configurable response period, verification, manual and automatic dispatch, GPS updates, response statuses, reports, and separately authorized announcements within a centralized workflow.

Economic Value: Builds on the barangay’s existing CCTV, network, and public-address infrastructure instead of replacing functioning systems. The prototype also allows evaluation before wider deployment and investment.

Competitive Value: Combines AI-assisted detection, three-level threat recommendations, authorized manual response, threat-based automatic dispatch for unattended alerts, GPS-supported responder coordination, structured reporting, and localized emergency announcements in a modular system designed for barangay operations.

Chapter 4. Startup Opportunity Analysis

4.1 Startup Readiness Assessment

Rating Scale

Table 2. Assessment Rating Scale

Assessment Criteria

Table 3. Startup Readiness Assessment Criteria

Individual Assessment Forms

Evaluator 1: Andam, John Carlo D.

Table 4. Individual Assessment Form - Andam, John Carlo D.

Evaluator 2: Baconawa, Cyrus Jeshurun Jacob M.

Table 5. Individual Assessment Form - Baconawa, Cyrus Jeshurun Jacob M.

Evaluator 3: Dela Pena, Merl Angelo S.

Table 6. Individual Assessment Form - Dela Pena, Merl Angelo S.

Evaluator 4: Ilio, Benedict L.

Table 7. Individual Assessment Form - Ilio, Benedict L.

Evaluator 5: Sanchez, Denver D.

Table 8. Individual Assessment Form - Sanchez, Denver D.

Group Summary

Table 9. Startup Readiness Assessment - Group Summary

Rating Interpretation

Table 10. Group Readiness Rating Summary

Analysis

The group evaluation produced an overall readiness score of 4.56, placing SafeWatch in the Highly Ready category. Problem-Solution Fit received the highest rating of 4.80, reflecting how the project addresses the challenges of monitoring approximately 200 CCTV feeds and coordinating incident response.

Innovation and Team Capability both received 4.60. SafeWatch combines AI-assisted detection, authorized manual decision-making, time-based automatic dispatch for unattended alerts, GPS-supported responder coordination, structured reporting, and localized fire announcements. Market Demand and Technical Feasibility both received 4.40, indicating strong potential while recognizing concerns involving AI accuracy, camera compatibility, processing capacity, network reliability, budget, and development time.

The project is supported by the barangay’s existing CCTV and network infrastructure, defined user roles, phased scope, and human-authorization controls. Before wider deployment, the team must validate selected camera feeds, evaluate AI performance, test role permissions and GPS functions, and gather formal user feedback during prototype acceptance.

4.2 Market and Competitor Analysis

Target Market

Primary Customers: Local government units, including barangays and city public-safety offices, and gated communities that operate or plan to deploy CCTV systems and require coordinated monitoring, incident verification, responder dispatch, and reporting.

Secondary Customers: Campuses, industrial estates, and other organizations that manage security personnel and need an integrated system for AI-assisted monitoring, GPS-supported response, controlled incident response, and incident documentation.

Competitor Analysis

Table 11. Competitor Analysis

Market Opportunity

SafeWatch addresses the needs of organizations that manage surveillance systems and field responders but still depend on manual monitoring or separate tools for incident verification, dispatch, tracking, reporting, and emergency communication. Potential customers include local government units, gated communities, campuses, industrial estates, and other organizations with defined security and response procedures.

The modular structure allows the system to be adapted to different user roles, incident categories, response procedures, camera deployments, and announcement zones. SafeWatch may be introduced as part of a new security setup or configured to work with compatible systems already in use. Its market opportunity is to provide an integrated detection-to-response platform combining AI-assisted monitoring, Command Center review, threat-based automatic dispatch for unattended alerts, GPS-supported coordination, and documentation.

Successful prototype validation may support further customization and deployment across different markets, subject to each client’s operational requirements, technical capacity, budget, procurement process, and data-privacy policies.

4.3 Technology Trends

Edge AI and Computer Vision: Modern object-detection frameworks such as YOLOv26 can analyze video locally, reducing dependence on high-latency cloud processing and allowing CCTV camera feeds to generate near-real-time alerts.

Multimodal Video: Visual evidence can provide additional context for configured incidents, while detection thresholds, noticed-alert review, automatic-dispatch timing, and post-incident records must be calibrated and evaluated during testing.

Human-in-the-Loop Artificial Intelligence: SafeWatch separates AI recommendations from human verification and public-announcement authorization while using an approved, rule-based automatic-dispatch function only when an alert remains unattended through the configured period.

Mobile GPS and Real-Time Notifications: Smartphone location services, maps, and push notifications support responder availability, proximity recommendations, navigation, and status monitoring.

API-First and Modular Integration: Application programming interfaces allow new monitoring and response functions to connect with selected data from existing systems without forcing the organization to replace all current infrastructure.

Records and Logs: Centralized time-stamped records improve accountability by linking AI detections, assignments, responder locations, and reports.

4.4 Unique Selling Proposition (USP)

USP Statement

"SafeWatch provides an AI-enabled, audio-assisted CCTV monitoring and incident-response platform for barangays by combining AI incident detection, threat-level recommendations, authorized manual response, time-based automatic dispatch for unattended alerts, GPS-enabled coordination, reporting, and localized announcements while preserving existing infrastructure and separate human authorization for public communication."

Key Competitive Advantages

Controlled Manual and Automatic Workflow: Authorized personnel may act during the configurable period; unattended Level 1, Level 2, and Level 3 detections

Complete Detection-to-Report Lifecycle: The system connects alert review, configured-period processing, threat classification, manual or automatic assignment, GPS monitoring, and dispatch reports.

Barangay-Specific Response Coordination: SafeWatch supports on-duty responder availability, proximity-based recommendations, approved threat-level routing, and mandatory mobile reporting.

Localized Announcement Audio: Confirmed fire incidents can produce selected-zone prerecorded announcements only after separate Command Center authorization.

Existing-Infrastructure-First Design: The project enhances current CCTV operations rather than requiring a complete replacement.

Chapter 5. Customer Validation

5.1 Customer Personas

Figure 4. Customer Persona - Sir Jolo Torreno

Figure 5. Customer Persona - Ma'am Leila Versoza

Figure 6. Customer Persona - Sir Joemar Lagarto

5.2 Interview Results

Validation interviews were conducted with a minimum of five (5) respondents representing key stakeholder roles at Barangay Holy Spirit Command Center, local business establishments, and community residents. The responses are summarized below:

Table 12. Interview Summary Table

5.3 Validation Findings

Key Positive Findings

Operational Alignment: SafeWatch follows the barangay’s existing chain of command by keeping incident verification, responder dispatch, and public announcements under authorized human control.

Improved Accountability: GPS-supported monitoring, structured incident statements, formal reports, and centralized records provide clearer documentation of incident response activities.

Infrastructure Compatibility: The modular design allows SafeWatch to use existing CCTV, network, and public-address infrastructure without requiring complete system replacement.

Common Concerns

AI Accuracy: False positives and missed detections may occur because of lighting, camera angles, crowding, weather, obstruction, and other environmental conditions.

Connectivity and Performance: Network interruptions, GPS availability, camera compatibility, and limited processing capacity may affect system performance.

System Usability: The PC and mobile interfaces must remain clear and responsive during incident review, dispatch, reporting, and other time-sensitive activities.

Overall Assessment

The validation process confirmed high overall interest and strong market demand among target users. Stakeholders validated that SafeWatch addresses real operational bottlenecks in continuous CCTV monitoring and fragmented response tracking. The core concept combining AI threat recommendations, human verification, GPS dispatch, and localized public announcements was enthusiastically received, provided that mobile connectivity issues and AI false-positive rates are systematically mitigated during prototype testing.

5.4 Recommended Improvements

Based on identified risks and planned prototype validation, the following improvements are recommended:

Connectivity and Data Recovery: Improve mobile connectivity handling through automatic reconnection and reliable data synchronization to reduce disruption to GPS updates, incident statements, and photo submissions during network interruptions.

Simplified Incident Review Interface: Keep the Command Center Personnel interface clear and responsive, with readable threat indicators, and efficient controls for confirming, rejecting, and monitoring AI detections and dispatches.

Optimized Selected-Feed Processing: Prioritize selected CCTV feeds and optimize AI processing to reduce server and network load before expanding to additional cameras.

Chapter 6. Lean Canvas

6.1 Problem

Continuous manual observation across approximately 200 CCTV cameras creates operator fatigue and makes it difficult to maintain equal attention across feeds, especially during simultaneous incidents, shift transitions, and busy periods.

Incident verification, responder coordination, location tracking, status updates, and reporting may be handled through separate processes, reducing efficiency and traceability.

Emergency announcements must be controlled and verified so that unverified AI detections do not trigger unnecessary public broadcasts.

An AI detection may remain unattended while Command Center personnel handle simultaneous monitoring, communication, and coordination duties, delaying an appropriate responder dispatch.

Validation basis: the Command Center Head identified monitoring fatigue; the Barangay Tanod identified unclear location details in phone/radio dispatch; the IT Department respondent emphasized reliability, access security, logs, and bandwidth; the business owner and resident supported faster coordination while retaining human control.

6.2 Customer Segments

Primary customer: Local government units, including barangays and city public-safety offices, with CCTV networks and defined response procedures.

Secondary customer: Gated communities, campuses, industrial estates, and other organizations managing security personnel and incident response.

Primary users: Command Center Personnel, IT Department Administrators, Purok Leaders, Barangay Tanods, and Barangay Task Force members.

Early-adopter profile: Organizations with existing CCTV and public-safety response workflows that still rely on manual monitoring or separate coordination tools.

6.3 Unique Value Proposition

“A controlled AI-assisted detection-to-response workflow that helps public-safety teams see possible incidents sooner, act within a configurable period, automatically dispatch appropriate responders when alerts remain unattended, document the full incident lifecycle, and issue separately authorized localized fire announcements without replacing functioning CCTV and public-address infrastructure.”

This value proposition preserves the approved balance: authorized personnel retain control over verification, final threat classification, manual actions, and public communication, while unattended detections follow the configured period and threat-level automatic-dispatch mapping.

6.4 Solution

Selected-feed AI detection with Level 1 Advisory, Level 2 Priority, and Level 3 Critical recommendations.

Command Center notification and a configurable period for review, verification, forwarding, or manual dispatch action.

Time-based automatic dispatch when an AI detection is not noticed within the configured period: the nearest and appropriate Barangay Tanods for Level 1 and Level 2, and the nearest and appropriate Barangay Task Force members for Level 3.

GPS-supported responder coordination, mobile status updates, incident statements, and formal reporting according to the assigned response workflow.

Central incident timeline plus separately verified and authorized localized announcements for confirmed fire incidents.

6.5 Channels

Direct institutional outreach to barangays, LGU public-safety offices, and other target organizations.

Live product demonstrations using the end-to-end detection-to-report workflow.

Professional/community networks, email outreach, referrals, and a product information page (proposed acquisition channels).

Deployment through configured PC and mobile applications connected to compatible CCTV, network, backend, GPS, and public-address resources.

6.6 Revenue Streams

Projected customer revenue consists of a ₱25,000 monthly platform subscription per active deployment and a ₱50,000 one-time deployment/configuration fee per new institutional customer. These rates are planning assumptions. The ₱3,604,624.21 initial project budget and ₱1,580,000 in estimated annual operational benefits are evaluated separately; neither is counted as customer revenue.

6.7 Cost Structure

Labor: ₱627,902.00, including the 10% contingency allowance in the initial project budget.

Hardware: ₱2,863,945.37, including the 10% contingency allowance in the initial project budget.

Software and services: ₱112,776.84, including the 10% contingency allowance in the initial project budget.

Total initial project cost baseline: ₱3,604,624.21. This is not a monthly operating cost or a selling price.

For operating projections, fixed costs are assumed at ₱30,000 per month and variable costs at ₱5,000 per active customer per month. These allowances exclude equipment and installation for additional deployments.

Installation, training, maintenance, technical support, and software updates are treated as separately scoped post-implementation services and are excluded from the operating projection.

6.8 Key Metrics

Approved initial project cost, actual expenditures, and contingency utilization.

Qualified institutional inquiries, demonstrations, pilot evaluations, paying customers, and continued use under an approved customer-payment model.

Incident workflow completion rate during testing: detection → configured-period result → manual or automatic dispatch → reporting.

AI false-positive/missed-detection rate and median time from alert to verified incident during controlled tests.

Estimated annual tangible benefits, cumulative benefits, ROI, and payback-period performance, measured separately from customer revenue.

Table 13. SafeWatch Lean Canvas

6.9 Unfair Advantage

SafeWatch is designed around a barangay-specific detection-to-report workflow that combines AI-assisted detection, three-level threat recommendations, authorized manual response, automatic dispatch for unattended alerts, responder GPS coordination, mandatory evidence and report submission, and separately authorized localized announcements while preserving existing infrastructure. Its long-term advantage will depend on deployment learning, stakeholder relationships, and accumulated workflow knowledge.

Chapter 7. Business Model

7.1 Business Model Overview

Obsidian IT Solutions will use a business-to-institution model centered on SafeWatch as a configurable public-safety platform. The initial customer focus is on LGUs and similar organizations that operate CCTV networks and need a more structured workflow for detection, verification, manual and automatic responder coordination, reporting, and controlled emergency communication. SafeWatch creates value by adding an AI-enabled application layer to the approved CCTV, network, processing, storage, GPS, and public-address environment.

The financial model assumes a ₱25,000 monthly platform subscription and a ₱50,000 one-time deployment/configuration fee per institutional customer. The ₱3,604,624.21 Barangay Holy Spirit implementation budget is analyzed separately from customer receipts. Estimated annual tangible benefits of ₱1,580,000 describe potential operational value, not sales revenue. Installation, training, preventive maintenance, technical support, and software updates are separately scoped services.

7.2 Value Proposition

SafeWatch helps public-safety organizations reduce the burden of continuously watching many CCTV feeds and reduce fragmentation between incident detection, verification, dispatch, tracking, reporting, and emergency communication. Its distinguishing value is not AI detection alone; it is the controlled end-to-end workflow in which AI recommends, authorized personnel may act during the configured period, unattended detections follow approved automatic-dispatch rules, responders receive actionable information, and incident evidence is recorded centrally. Public announcements remain separately verified and authorized.

7.3 Revenue Model

The revenue model combines a monthly platform subscription for continuing institutional use with a one-time deployment/configuration fee for each new customer. These are projected customer payments rather than contracted sales. The ₱3,604,624.21 initial implementation budget is treated separately from recurring platform operations.

Table 14. Revenue Model

Estimated operational benefits are not a revenue model or customer payment.

7.4 Customer Relationship

The proposed customer lifecycle is direct and service-supported: identify qualified institutions → conduct discovery and demonstration → confirm operational requirements and compatibility → configure and onboard authorized users → provide initial orientation and support → monitor issues and feedback → review performance and renewal needs. Customer feedback should be collected through structured check-ins and post-demo/pilot feedback, with technical issues routed into the development backlog. Retention depends on dependable operation, clear user workflows, responsive support, and demonstrated value in incident monitoring and documentation.

7.5 Key Partners and Resources

Table 15. Key Partners and Resources

Chapter 8. Minimum Viable Product (MVP)

8.1 MVP Description

The SafeWatch MVP is planned as the smallest testable version of the detection-to-response workflow: selected CCTV feed analysis → AI alert and threat recommendation → Command Center notice and review within the configured period, or automatic dispatch when the alert remains unnoticed → responder GPS and status updates → applicable incident statement or formal report → centralized incident record. Under the unattended-alert workflow, Level 1 and Level 2 detections dispatch the nearest and appropriate Barangay Tanods, while Level 3 detections dispatch the nearest and appropriate Barangay Task Force members. A confirmed fire announcement remains separate from responder dispatch and requires Command Center verification and authorization.

These planned functions address monitoring fatigue, unclear dispatch locations, and fragmented records identified during customer validation. MVP testing will evaluate end-to-end operation and role controls before deployment.

8.2 Features Included

Table 16 defines the planned MVP functions and links each function to an operational need.

Table 16. MVP Features Included

8.3 Product Roadmap

Table 17. Product Roadmap

8.4 Screenshots / Prototype

Figure 3 in Chapter 3 depicts the SafeWatch physical prototype assembly. It illustrates the CCTV and public-address hardware concept, while the PC, mobile, AI, and backend workflows described above remain the scope of MVP implementation and testing.

Chapter 9. Marketing Plan

Marketing for SafeWatch should be treated as institutional outreach rather than mass-market consumer advertising. Potential customers include LGUs, barangays, city public-safety offices, gated communities, campuses, industrial estates, and other organizations with defined security and response procedures. Adoption will depend on operational need, budget, approval, compatibility, procurement, and data-privacy requirements.

The initial marketing objective is to demonstrate the complete SafeWatch workflow, support controlled pilot evaluation, document stakeholder feedback, and use verified operational results to guide future implementation discussions.

9.1 Target Market

Table 18. SafeWatch Target-Market Profile

9.2 Market Segmentation

Table 19. Market Segmentation

9.3 Pricing Strategy

The ₱3,604,624.21 initial project cost baseline includes labor, hardware, software and services, and 10% contingency. For operating projections, SafeWatch assumes a ₱25,000 monthly platform subscription, ₱50,000 one-time deployment/configuration fee, ₱5,000 monthly variable cost per active customer, and ₱30,000 monthly fixed cost. A separate cost-plus illustration prices one complete implementation at ₱3,965,086.63, or 10% above the project cost baseline; it is excluded from the subscription forecast. These figures are planning assumptions, not customer agreements.

Table 20. Pricing Strategy

At the assumed subscription rate, contribution before fixed costs is ₱25,000 − ₱5,000 = ₱20,000 per active customer per month. With two active customers, ₱50,000 in recurring revenue less ₱40,000 in modeled fixed and variable costs leaves a ₱10,000 monthly operating surplus. The calculation excludes initial implementation and customer-specific equipment and installation costs, so it does not represent full deployment profit.

9.4 Promotion Strategy

Table 21. Promotion Strategy

9.5 Distribution Channels

SafeWatch will be delivered as a configured system implementation that combines the PC and mobile applications with the approved processing, storage, camera, networking, backend, mapping, and public-address resources. After deployment approval, Obsidian IT Solutions will configure authorized accounts and roles, connect compatible CCTV camera feeds, configure incident categories, the automatic-dispatch period, responder routing, reporting, and speaker zones, and orient authorized users. Command Center and IT Administration functions are accessed through PC applications, while Purok Leader, Barangay Tanod, and Barangay Task Force functions are accessed through mobile applications. The implementation uses Windows Server 2025 Standard, Supabase Pro, YOLOv26, and OpenStreetMap API together with the approved hardware environment.

Chapter 10. Operations Plan

10.1 Organizational Structure

The startup has five members with the functional assignments and named members shown in Table 23. The Project / Business Lead coordinates project direction, business planning, milestones, stakeholder communication, budget, and approvals. The Technical / AI Lead, Backend / Integration Lead, Mobile / QA Lead, and Marketing / Finance / Operations Lead carry out the functions described in the table.

Table 22. Proposed SafeWatch Organizational Structure

Table 23. Organizational Structure and Functional Roles

10.2 Team Roles

The following responsibilities correspond to the five functional roles listed in Section 10.1.

Table 24. Team Roles

10.3 Development Workflow

SafeWatch should use an iterative development workflow in which validated customer needs are converted into prioritized work, developed in small increments, tested against the documented risks, and returned to the backlog for improvement. Planned validation focuses on authentication, permissions, AI detection, incident verification, configured-period processing, manual and automatic dispatch, GPS, notifications, status changes, reporting, and separate public-announcement authorization.

Table 25. Development Workflow

10.4 Resources Required

Table 26. Resources Required

10.5 Risk Management

Table 27. Risk Management

The likelihood ratings are qualitative planning judgments informed by customer feedback and the technical risks identified in Chapters 4 and 5. The principal risks concern AI accuracy, connectivity and performance, camera compatibility, processing capacity, and usability.

Chapter 11. Financial Assumptions and Evidence

The financial analysis separates the ₱3,604,624.21 Barangay Holy Spirit initial project budget, ₱1,580,000 in estimated annual operational benefits, and projected institutional customer payments. Operational benefits are not sales receipts. Tables 28–34 apply explicit assumptions to illustrate subscription revenue, operating costs, cash flow, and scenarios.

Table 28. Financial Assumptions and Evidence

Sections 11.2–11.7 use the subscription-and-fee model for a 12-month operating projection. Initial implementation investment, equipment and installation for additional customer sites, taxes, and financing are outside this operating model. Project investment recovery is assessed separately using the estimated annual operational benefits.

11.1 Startup Cost

Table 29 presents PROJMAN’s initial resource budget for a SafeWatch implementation in Barangay Holy Spirit. For this analysis, the ₱3,604,624.21 total serves as the initial project investment, separate from monthly platform operating costs.

Table 29. Startup Cost

Total Initial Project Cost = ₱627,902.00 + ₱2,863,945.37 + ₱112,776.84 = ₱3,604,624.21. The category totals include 10% contingency and agree arithmetically. The software subtotal of ₱102,524.40 implies ₱18,824.40 for 12 months of Supabase Pro at ₱1,568.70 per month, plus the ₱83,700.00 Windows Server license; the printed Supabase line of ₱4,714.50 is inconsistent with that 12-month term. The analysis uses PROJMAN’s repeated ₱3,604,624.21 cost baseline.

The labor allocation covers the IT Project Manager, Full-Stack Web and UI/UX Developer, Junior QA Tester, Hardware Engineer, and Software Engineer for the planned project period. The hardware allocation covers processing components, an AVR, PoE switch, hard drives, NVRs, 200 IP cameras, and 50 public-address units. The software and services allocation covers Windows Server 2025 Standard, Supabase Pro, YOLOv26, and OpenStreetMap API. The 10% contingency is already included and must not be added again.

For planning, ₱3,604,624.21 is the startup investment benchmark for the initial implementation. The operating cash projection in Section 11.6 begins after this investment and therefore does not include the initial project outlay.

11.2 Revenue Streams

Table 30 applies the assumed subscription and deployment/configuration prices to institutional customers. The figures describe projected customer payments and exclude the separate complete-implementation price.

Table 30. Revenue Streams

Under the 12-month acquisition scenario, 36 active-customer months generate ₱900,000 in subscription revenue and four deployment/configuration fees generate ₱200,000, for ₱1,100,000 in projected receipts. These are forecasts rather than executed sales. The ₱1,580,000 in estimated annual operational benefits is excluded from revenue.

11.3 Cost Structure

The ₱3,604,624.21 initial implementation budget is a one-time project investment. The monthly model assumes fixed operating costs of ₱30,000 and variable costs of ₱5,000 per active customer. It excludes equipment, installation, and other site-specific costs for additional deployments; accordingly, its operating results are not full project profit.

Table 31. Cost Structure

At two active customers, modeled monthly operating cost is ₱30,000 + (2 × ₱5,000) = ₱40,000. Hardware is the largest component of the initial implementation budget. The recurring allowance rises by ₱5,000 per additional customer within the assumed range, while separately scoped support, renewals, replacement, installation, and training are outside this operating illustration.

11.4 Sales Forecast

The 12-month forecast assumes ₱25,000 per active customer per month, ₱50,000 per new customer, ₱5,000 per active customer per month in variable cost, and ₱30,000 per month in fixed cost. It models one new institutional customer in Months 1, 3, 5, and 7. This is an illustrative acquisition projection rather than a record of contracts. The separate ₱3,965,086.63 complete-implementation price is excluded.

Table 32. Sales Forecast

Projected Year 1 receipts comprise ₱900,000 in subscriptions and ₱200,000 in deployment/configuration fees. Under the assumed operating allowances, Month 2 shows a ₱10,000 deficit and Month 3 onward shows positive monthly surpluses. The ₱560,000 Year 1 operating surplus precedes the ₱3,604,624.21 initial investment and any additional deployment costs, so it is not net project profit.

11.5 Break-even Analysis

Using the recurring operating assumptions:

Contribution Margin per Active Customer = ₱25,000 − ₱5,000 = ₱20,000 per month.

Operating Break-even Quantity = ₱30,000 ÷ ₱20,000 = 1.5 customers. Rounded up, the conditional operating break-even point is 2 active paying customers per month.

Operating Break-even Sales Revenue = 2 × ₱25,000 = ₱50,000 in recurring monthly revenue, excluding one-time deployment/configuration fees.

At two customers, ₱50,000 in recurring revenue less ₱10,000 in variable costs and ₱30,000 in fixed costs leaves a ₱10,000 monthly modeled surplus; two customers are needed because the algebraic threshold is 1.5. The acquisition projection reaches that level in Month 3. This operating break-even excludes initial investment and additional deployment costs. Separately, PROJMAN’s preliminary gross-benefit calculation implies 2.28-year project payback and 31.50% three-year ROI before recurring operating costs, if the ₱1,580,000 estimated annual benefits recur.

11.6 Cash Flow Projection

The operating cash projection assumes ₱30,000 opening cash, equivalent to one month of modeled fixed costs, and same-month collection and payment. It begins after initial implementation and includes only subscription and deployment/configuration receipts and monthly operating allowances. It excludes initial project investment, additional site costs, taxes, and financing flows.

Table 33. Cash Flow Projection

With these assumptions, net cash flow totals ₱560,000 and ending operating cash reaches ₱590,000 after 12 months. Month 2 shows a ₱10,000 net outflow because no deployment/configuration fee is received. Delayed collections or expenses outside the modeled allowances would reduce the projected balance.

11.7 Scenario Analysis

The scenarios vary active paying customers while holding the assumed monthly subscription of ₱25,000, variable allowance of ₱5,000 per customer, and fixed allowance of ₱30,000 constant. They exclude one-time deployment/configuration fees and initial project investment.

Table 34. Scenario Analysis

Each additional active customer adds ₱20,000 in modeled monthly contribution before fixed costs. One customer produces a monthly operating deficit, while two and four produce surpluses under the stated allowances. The two-customer Expected scenario represents an early operating benchmark; the four-customer Optimistic scenario corresponds to Months 7–12 of the acquisition forecast. Results remain sensitive to actual pricing, collections, and site-specific costs.

References

[1] Milestone Systems A/S, “XProtect Video Management Software.” [Online]. Available: https://www.milestonesys.com/products/software/xprotect/. Accessed: Aug. 31, 2026.

[2] Genetec Inc., “Mission Control Incident Management System.” [Online]. Available: https://www.genetec.com/products/operations/mission-control. Accessed: Aug. 31, 2026.

[3] Avigilon, “Unity On-Premise Video Security.” [Online]. Available: https://www.avigilon.com/vms/on-premise. Accessed: Aug. 31, 2026.

[4] Hangzhou Hikvision Digital Technology Co., Ltd., “HikCentral Professional Integration.” [Online]. Available: https://tpp.hikvision.com/products/HCP-Integration. Accessed: Aug. 31, 2026.