Consultation procedures
A process that lets employees seek help and safely reach the responsible team.
HUMAN PROTECTION AI / PATENT & DEVELOPMENT
Not to monitor people.
Not to monitor people.
To protect them.
Published patented detection and recording technology, alongside commissioned development of cameras, video analytics, Edge AI and management platforms. These two pathways support discussions of site-specific PoCs and joint development.
Design consultation, fact-finding and organizational responses, not just equipment-based compliance.
In Japan, from October 1, 2026, all employers must take employment-management measures to prevent workplace customer harassment.
Fact-finding
Not every customer complaint is harassment. Legitimate requests within socially acceptable bounds are not harassment; people must review the individual circumstances.
Organizational response
Appropriate recording methods, protection of those involved and prevention of recurrence.
AI or audio/video recording equipment alone does not complete legal compliance. AI supports record organization and review; people and organizations make the final decisions.
Amended law (Japanese)Commencement order (Japanese)Ministry guidelines (Japanese)Implementation notice (Japanese)
Discuss a records and operations PoC
PATENTPublished detection and recording technology
DEVELOPMENTCameras, video analytics, Edge AI and management platforms
PoCPhased validation for site conditions
PARTNERSHIPIntegration, joint development and business partnerships
Choose a starting point for your inquiry.
01
Field deployment and employee protection
For organizations considering fact-finding, sharing information with managers and field PoCs.
Explore use cases
02
Integration into products and devices
For businesses considering AI integration into recording devices, licensing and joint development.
Explore patents and integration
03
AI camera and platform development
For organizations considering camera development, video analytics, Edge AI, management interfaces and API integration.
Explore development areasRead the full image translation beside the larger image below
Customer harassment prevention: beyond compliance, so frontline staff are not left alone.
In Japan, customer-harassment prevention measures become mandatory for employers on October 1, 2026 under Act No. 63 of 2025, promulgated June 11, 2025, amending the Act on Comprehensive Promotion of Labor Policies and related laws. The prevention guidelines are Ministry of Health, Labour and Welfare Notice No. 51 of 2026, issued February 26, 2026.
of companies reported an increase in consultations known to them
of companies had taken no customer-harassment measures
of workers reported no particular initiatives at their workplace
Source: FY2023 Survey Report on Workplace Harassment, commissioned by Japan's Ministry of Health, Labour and Welfare (March 2024, PwC Consulting LLC). Valid company responses: 7,780; worker and related respondents: 8,000.
AI helps make records more objective; it does not decide.
What AI supports
- Using audio and images as material for review
- Supporting records from before detection through the end of the conduct
- Highlighting intervals for the responsible person to review
- Sharing factual information within the organization
What AI does not replace
- Determining illegality or criminal conduct
- Disciplinary or legal decisions
- Completing statutory measures through equipment alone
- Staffing and consultation arrangements
Image text translation
Reference translation prepared for this website. The source image is unchanged; this is not an original or certified translation.
Safeguard Review / incident review details / ID: IR-2024-0715-001
- Privacy-first mode / help / auditor
- Dashboard / incident list / incident review / evidence management / access management / reports / settings / audit log / privacy settings / close menu
Privacy protection active / all personal information is anonymized and minimized / view details
Basic incident information
- Type: suspected harassment / occurred: 2024/07/10, 15:30–16:05 / location: internal meeting room / reporting channel: consultation desk / status: under review / updated: 2024/07/15, 10:42
Privacy settings / access control
- Anonymization (names, voices and images) / hide affiliation information / third-party sharing: restricted
- Roles permitted to view this incident: compliance staff and auditors / manage access permissions
Recorded data (still image) / data information
- Playback: 00:12:34 / 00:35:21 / show full screen
- Recording: VID_20240710_153005.mp4 / size: 1.2 GB / resolution and frame rate: 1920×1080 / 30 fps
- Capture device: meeting-room camera (Room-3) / storage: encrypted evidence repository / SHA-256 sample: a1b2c3d4...e9f8a7b6 / view evidence details
Processed transcript (anonymized) / show AI summary
- 00:03:15 Speaker A: You would not normally understand that, would you? Why can you not do it, no matter how many times I tell you?
- 00:03:24 Speaker B: I am sorry. I will be more careful next time.
- 00:03:31 Speaker A: Be careful? It is beyond that. You made the same mistake before, did you not?
- 00:03:45 Speaker A: Everyone else is doing it properly. You are the only one. Do you really want to do the work?
- 00:04:02 Speaker B: ... / personally identifying information has been anonymized / show full text
AI assessment summary / AI analysis score: high risk, 78/100
The screen's sample assessment states that the statements and context indicate a high possibility of harassment.
- Intimidating language: high / personal denigration and comparison: high / repetition and persistence: medium / psychological impact: high
The score is an AI-assisted assessment; a human review makes the final determination. / View AI assessment details
Detected risk items
- Personal denigration: 'You are the only one', high, 00:03:45 / excessive reprimand: 'Can you not do it, no matter how many times I tell you?', high, 00:03:15
- Revisiting past mistakes: 'You made the same mistake before', medium, 00:03:31 / questioning motivation: 'Do you really want to do the work?', high, 00:03:45 / show all detections
Human review and comments
- Reviewer: compliance staff / review time: 2024/07/15, 10:30 / comments
I reviewed the AI assessment and detected items. Considering the tone, context and continuity of the statements together, I judge that they may be likely to constitute workplace power harassment. Please proceed with an interview of the person concerned and confirmation with the parties involved.
- Edit comments
Storage and retention settings
- Retention policy: standard policy (internal rules) / retention: 5 years / planned deletion: 2029/07/15 / change (three controls)
After the retention period, the screen states that data will be securely deleted in accordance with policy.
- Complete review / request further investigation / close incident
Video and evidence metadata are on the left, anonymized transcript and AI-assisted assessment in the center, and risk items, human review and retention settings on the right. Bottom controls complete, investigate or close the review.
Site note: this is a conceptual screen. Scores, dates, findings and retention settings are examples, not a legal determination, validated accuracy, guaranteed anonymization or a prescribed retention period.
Enlarge patent publication
Image text translation
Reference translation prepared for this website. The source image is unchanged; this is not an original or certified translation.
JP 7728631 B1, 2025.8.25 / (19) Japan Patent Office (JP) / (12) Patent Gazette (B1)
- (11) Patent number: Japanese Patent No. 7728631 (P7728631)
- (45) Publication date: August 25, Reiwa 7 (2025.8.25) / (24) Registration date: August 15, Reiwa 7 (2025.8.15)
- (51) International Classification: G10L 15/10 (2006.01) / FI: G10L 15/10 500Z / Number of claims: 3; total pages: 15
- (21) Application: Japanese Patent Application 2025-76317 (P2025-76317) / (22) Filing date: May 1, Reiwa 7 (2025.5.1) / Examination requested: May 1, Reiwa 7 (2025.5.1) / Application subject to accelerated examination
- (73) Patentee: 524337664, Standard Link Co., Ltd. Address as printed in the gazette: TS Building, 2nd floor, 2-9-9 Iwamotocho, Chiyoda-ku, Tokyo.
- (74) Representative: 100166589, patent attorney 植村貴昭 (name retained as printed)
- (72) Inventor: 大西俊幸 (Toshiyuki Onishi), at Standard Link Co., Ltd., TS Building, 2nd floor, 2-9-9 Iwamotocho, Chiyoda-ku, Tokyo. Examiner: 本多美優梨 (name retained as printed). Continued on the final page.
(54) Title of invention: Harassment detection and recording device
(57) Claims / Claim 1
A sensor that collects audio data and image data by capturing sound and images from the position at which it is mounted;
detection means that detects harassment behavior based on, and by combining, the audio data and image data collected by the sensor; and
control means that, when harassment behavior is detected by the detection means, stores in a storage device, from among the audio data and image data collected by the sensor, at least the audio and image data from before detection of the harassment behavior until the harassment behavior ends after detection;
a harassment detection and recording device comprising the above and worn on clothing.
Claim 2
Further comprising temporary storage means that temporarily stores the audio data and image data collected by the sensor;
the sensor collecting the audio data and image data continuously in real time, or collecting them by periodic sampling;
the control means, when harassment behavior is detected by the detection means, storing in the storage device, from among the audio data and image data stored in the temporary storage means, at least the audio and image data from before detection of the harassment behavior until the harassment behavior ends after detection,
The upper half contains bibliographic information. The lower half shows Claim 1 and the opening of Claim 2; Claim 2 continues beyond this image. The right-margin numbers 10 and 20 are line numbers. This image is page 1, not all 15 pages or all three claims.
Site note: this reference translation does not determine legal claim scope. The Japanese gazette is the source document, and the address is reproduced as published at that time, not as the current company contact address. No omitted continuation has been invented.
Place intellectual property at the heart of the business.
The granted publication discloses configurations combining audio and images and preserving records before and after detection. Refer to the granted publication for the precise scope of rights.
Published documentsThis explanation does not limit rights to a particular installation method or device shape
Audio + imagesCombination of information described in the granted publication
Record preservationConfiguration covering periods before and after detection
Image text translation
Reference translation prepared for this website. The source image is unchanged; this is not an original or certified translation.
How the harassment detection and recording device works
- Sample badge label: 山田太郎
Camera
Captures facial expressions and movements for behavior analysis.
Microphone
Collects audio for analysis of tone and volume.
- Data collection: acquire audio and video in real time
- AI analysis and detection: detect harassment using multiple algorithms
- Automatic recording: automatically retain data from before and after detection
- Warnings and notifications: vibration alerts and notifications as needed
Leader lines locate the camera and microphone on the illustrated badge. The four lower boxes run left to right: collect, analyze / detect, record, then alert / notify.
Site note: this is a concept illustration, not the complete patent claims or a guarantee of a finished product, detection accuracy or every depicted function.
Device configuration example from published materials
Body camera
For in-person services and mobile work
Vertical device
Portable use at reception and information counters
Horizontal device
For discussions of compact, portable form factors
These examples are implementation candidates, not determinations of whether individual products fall within patent rights.
Patented technology and commissioned development: two clearly separate areas.
Refer to the granted publication for the precise scope of Japanese Patent No. 7728631. Fixed cameras, general video analytics, Edge AI and management platforms below are presented as SLL's separate commissioned-development business, not as an explanation of that patent's scope.
Separate explanations of patented technology and AI development
Elements disclosed in the granted publication
- Combining audio and images
- Preserving records before and after detection
- See the granted publication for details
This page does not determine whether individual products fall within patent rights.
Development areas offered as a separate business
- Cameras and video analytics
- Edge AI
- Management platform
- APIs and data infrastructure
Discuss commissioned development, PoCs and joint development.
Discuss AI developmentStandard Link distinguishes intellectual-property explanations from implementation services.
DEVELOPMENT / WHAT WE CAN BUILD
Turn existing cameras, video and audio into resources for incident response.
An application layer that analyzes information available on site, including existing surveillance cameras, uploaded video, audio and portable devices, supporting incident discovery, clipping, objective review, evidence organization and sharing.
Standard Link covers camera and device selection and development, AI analytics, edge processing, management interfaces and operational design as an integrated project.
Read the full image translation beside the larger image below
Find and extract
Make it easier to locate video and audio intervals that staff should review.
Organize records
Organize related video, audio and text as incident-level review material.
Manage and share
Connect review, sharing and response records according to permissions and operating procedures.
Image text translation
Reference translation prepared for this website. The source image is unchanged; this is not an original or certified translation.
Harassment AI management system / case list > case details
- Back to list (shown twice) / dashboard / case list / monitoring / reports / master-data management / settings / close menu
On-site video (camera 1) / recording
- Displayed video timestamp: 2025-05-20 10:24:17; playback label: 10:24:17 / 30:00.
Basic information
- Case ID: 2025-05-20-0001 / occurred: Tuesday, 2025/05/20, 10:23 / location: main-branch reception counter / status: under investigation
- Reporter: 佐藤花子 (reception staff) / reported: Tuesday, 2025/05/20, 10:30 / first responder: 鈴木誠 (deputy branch manager) / last updated: Tuesday, 2025/05/20, 11:15
- Tags: customer harassment / intimidating words or behavior / prolonged interaction
Case overview
Summary: an incident involving intimidating behavior and loudly voiced demands by a customer during an explanation of counter procedures.
- Department: counter services / parties: male walk-in customer; responding staff: 佐藤花子 (reception) / related file: reception-counter layout diagram (.pdf)
Transcript / show full text / time, speaker and spoken content
- 10:23:41 Customer: Enough already! How many times do I have to explain before you understand?
- 10:23:46 Staff: I am sorry. I will explain step by step, so please wait a moment.
- 10:23:53 Customer: I cannot wait! I am busy! Do something now!
- 10:24:02 Staff: Could you please refrain from raising your voice? It is disturbing other customers.
- 10:24:09 Customer: Shut up! I do not like your attitude!
- 10:24:15 Staff: Understood. I will contact the person in charge; please wait a moment.
- 10:24:21 Customer: Get the person in charge! Right now!
The transcript is automatically generated by AI; its accuracy is not guaranteed.
- Download transcript
The left panels pair video with case metadata; the right panels show the overview and time-stamped transcript. Playback, volume, full-screen and download icons indicate the proposed controls.
Site note: this is a conceptual development screen. Names, incidents and timestamps are sample screen content, not a report of a verified real case or a completed product.
Image text translation
Reference translation prepared for this website. The source image is unchanged; this is not an original or certified translation.
1. Using existing surveillance cameras
Analyze existing camera footage with AI in one workflow for discovering incidents, creating clips, supporting objective review and organizing evidence.
Existing surveillance cameras
Import footage from cameras already installed in shops and facilities into the platform.
- Secure import of existing camera footage / on-premises or cloud configurations
Harassment-response platform / incident list and details
- Dashboard / incident list / video search / clip list / reports / settings
- Scene information: 2024/05/15 14:32:18; camera: counter 2; location: first-floor general service counter; detection category: intimidating speech / shouting; AI score: 0.92 (high); notes: none
- Timeline labels: 14:20, 14:30, 14:40, 14:50 / AI-detected scenes
- AI-detected scenes (3): 14:32:18–14:33:02, high, intimidating speech / shouting, duration 00:00:44
- 14:35:47–14:36:21, medium, persistent / repeated demands, 00:00:34; 14:41:09–14:41:36, medium, pointing / intimidating attitude, 00:00:27
- Export clip / add to report
Toward faster response and recurrence prevention
Use evidence footage to speed up fact checking and support appropriate responses and recurrence prevention.
- Secure storage and permission management / manage evidence in a secure environment intended to support reliability
- Automatic AI detection: identifies scenes that may involve harassment
- Clip relevant scenes: automatically select clip lengths including preceding and following context
- Organize evidence: chronological records for each incident, with report output
- Support prompt action: the concept aims to substantially reduce fact-checking time and enable early response and recurrence prevention
Reuse existing cameras. The concept aims to make harassment response more dependable and faster with AI.
Arrows lead from existing cameras on the left to AI detection / clipping in the center and human review / retention on the right. Four functions are summarized below.
Site note: a development configuration example. Scores such as 0.92 are screen examples, not validated detection performance or a guarantee of legal evidentiary value. Compatibility and operational requirements must be assessed individually.
Turn existing video into manageable incident intervals
Receive video from cameras already installed in shops, stations, hospitals and public counters; extract intervals for review and organize clips with related information by incident. Define connection methods, recording formats and storage locations individually in the PoC.
Image text translation
Reference translation prepared for this website. The source image is unchanged; this is not an original or certified translation.
Pattern 2: uploaded-video analysis
Upload an existing video; the proposed AI workflow analyzes it and organizes and visualizes material for evidence review.
- Upload: drag and drop / analysis: AI analyzes video and audio / key-scene extraction: automatically detect important scenes / summarization: summarize content and key points / evidence organization: centralized management and sharing
- Drag and drop a video file here, or select a file
- Videos from different devices and settings: security camera (MP4), smartphone (MOV), meeting / interview recording (MP4); listed formats include MP4, MOV and AVI
Analyzing... 78%
- Audio extraction: audio extracted from video / transcription: audio converted to text
- Key-scene detection: 3 important scenes extracted; scene 1 at 00:02:15, scene 2 at 00:07:42, scene 3 at 00:12:31; summary generated from the video's key points
Summary
- This video records an event concerning ○○ on 2024/05/15. At the beginning, the parties exchange explanations and confirmations.
- Around 07:42, problematic statements and behavior are identified. The parties then discuss and agree on a response, finally deciding on measures intended to prevent recurrence.
- Key scenes: 3 / total duration: 15 minutes 32 seconds / transcription: complete
Incident-record details
- Share / export / overview; title: incident record concerning ○○; recorded: 2024/05/15, 10:00–10:15; location: meeting room A; parties: ○○, △△, □□; tags: harassment, meeting record
- Key scenes (3): 00:02:15 incident overview; 00:07:42 problematic statements / behavior; 00:12:31 response discussion and agreement; play / transcript / search / show full text
- Potential uses: internal incident / harassment investigations; reviewing customer-service records; recording and reviewing compliance training; supporting materials for audits / administrative responses
The concept supports compliance review by visualizing facts for fair, neutral judgment, managing evidence centrally for sharing among authorized parties, and aiming to reduce time spent on audits and investigations.
The left-side upload feeds into extraction, transcription, key-scene selection and summary generation in the center, then into a record for human review on the right.
Site note: a development-screen example, not validated accuracy, timing, universal format support, legal evidentiary value or a measured time-saving guarantee. The example's different duration labels are reproduced rather than recalculated.
Analyze recorded files later with AI
Post-process exported security-camera footage, smartphone videos and meeting or interview records. Rather than requiring real-time processing, focus on extracting important scenes, summarizing and generating review clips.
Image text translation
Reference translation prepared for this website. The source image is unchanged; this is not an original or certified translation.
3. Audio transcription and extraction of items requiring review
Receive audio, transcribe it and extract key points. Review it objectively and organize the record.
Even where cameras cannot be used, recorded audio can help visualize and analyze interactions, supporting service-quality improvement and risk management.
Receive different types of audio
- Headset audio: call-center or consultation-desk operator calls
- IC recorder audio: in-person consultations or visit recordings
- Call recordings: PBX or cloud-telephone audio / upload audio files; example formats: mp3, wav, m4a and others
Automatic transcription and key-point extraction
- Sample file: 2024_0515_相談対応_鈴木様.mp3 (consultation response, Suzuki); playback speed: 1.0×; create segment; playback 00:00 / 32:48; timeline 0:00, 05:00, 10:00, 15:00, 20:00, 25:00, 30:00
- Automatic speaker separation: speaker 1 (operator) / speaker 2 (customer)
- Tone and volume trends: tone from calm to tense; volume from low to high; graph times 00:00, 10:00, 20:00, 30:00
Automatic transcript (Japanese in the source screen)
- 00:01:12 Speaker 2: I received a bill that differs from the contract, and I cannot accept it. Review required.
- 00:01:28 Speaker 1: I am sorry for the inconvenience. Let me check the details.
- 00:02:45 Speaker 2: I thought I had canceled, but I am still being charged. Review required.
- 00:03:02 Speaker 1: Understood. Please let me confirm when you contacted us to cancel.
- 00:04:15 Speaker 2: I called in early April. ...
- Automatically extracted terms / phrases requiring review: differs from the contract; cannot accept it; thought I had canceled; still being charged; when cancellation was requested; ...
Objective review and organization of records
- Review interactions: use transcripts, speaker separation and tone trends to support accurate understanding of facts.
- Efficiently review flagged points: prioritize automatically extracted important passages to help avoid omissions and risks.
- Organize and share records using search, tags and notes, with secure evidence management.
- Search and filter / tags and notes / export / permission management
- Security / compliance design items: encrypted communication (SSL/TLS), encrypted data storage, access-permission management, operation logs and a domestic data center
Three input types on the left feed the central transcription / key-point process, then human review and record management on the right. The lower band lists security design items.
Site note: a concept screen, not verified transcription accuracy, speaker / emotion recognition, format compatibility, guaranteed security or a universal domestic-storage promise. These conditions require project-specific confirmation.
Review audio objectively after recording where cameras are absent
After recording, apply speaker separation, transcription and extraction of review intervals to call-center calls, telephone inquiries, interviews and voice-recorder files. Prioritize reliable retrospective review and stable operations rather than mandatory real-time processing.
Enlarge in a new window Image text translation
Reference translation prepared for this website. The source image is unchanged; this is not an original or certified translation.
4. Multi-camera integrated timeline
Automatically integrate footage from multiple cameras and optional wearable devices to manage one incident chronologically in a single place.
Various input sources
- 1. Entrance camera, 192.168.1.10 / 2. Aisle camera, 192.168.1.11 / 3. Checkout / counter camera, 192.168.1.12 / 4. Optional wearable camera, 10.0.0.55
- POS / audio / alarms / sensors and other inputs
Integrated timeline: chronological management as one incident
- Before (preparation / early signs) → occurrence (incident begins) → response (actions / handling) → end (resolution)
- Time labels: 14:22, 14:24, 14:26, 14:28, 14:30; selected time: 14:24:36; four synchronized rows: entrance, aisle, checkout / counter and optional wearable camera
- 1× playback / bookmark / create clip / export
An integrated evidence package to support timely, informed decisions
- Review all angles on the same timeline / help prevent missing evidence / output reports and supporting documents / support fair and objective judgment
- Synchronize footage: automatically combine video from different cameras and devices chronologically
- Organize as one incident: manage dispersed footage on a single timeline
- Access key moments: move along the timeline to inspect any chosen point quickly
- Reduce review time: streamline search, clipping and export, aiming to reduce operational effort
Each of four input arrows corresponds to one footage row. A red vertical marker aligns the selected instant across all rows. Human review is shown on the right; intended benefits are listed below.
Site note: IP addresses and times are illustrative labels in the concept image, not actual network access information or proven synchronization performance. Device integration, time alignment and evidence handling require individual validation.
Multiple records, one incident timeline
Align sources such as entrances, corridors, checkout counters, service desks and optional portable-device video on one timeline. Do not promise unconditional identity matching; validate incident organization using times, locations and event information in stages.
Image text translation
Reference translation prepared for this website. The source image is unchanged; this is not an original or certified translation.
4. Multi-camera integrated timeline
Automatically integrate footage from multiple cameras and optional wearable devices to manage one incident chronologically in a single place.
Various input sources
- 1. Entrance camera, 192.168.1.10 / 2. Aisle camera, 192.168.1.11 / 3. Checkout / counter camera, 192.168.1.12 / 4. Optional wearable camera, 10.0.0.55
- POS / audio / alarms / sensors and other inputs
Integrated timeline: chronological management as one incident
- Before (preparation / early signs) → occurrence (incident begins) → response (actions / handling) → end (resolution)
- Time labels: 14:22, 14:24, 14:26, 14:28, 14:30; selected time: 14:24:36; four synchronized rows: entrance, aisle, checkout / counter and optional wearable camera
- 1× playback / bookmark / create clip / export
An integrated evidence package to support timely, informed decisions
- Review all angles on the same timeline / help prevent missing evidence / output reports and supporting documents / support fair and objective judgment
- Synchronize footage: automatically combine video from different cameras and devices chronologically
- Organize as one incident: manage dispersed footage on a single timeline
- Access key moments: move along the timeline to inspect any chosen point quickly
- Reduce review time: streamline search, clipping and export, aiming to reduce operational effort
Each of four input arrows corresponds to one footage row. A red vertical marker aligns the selected instant across all rows. Human review is shown on the right; intended benefits are listed below.
Site note: IP addresses and times are illustrative labels in the concept image, not actual network access information or proven synchronization performance. Device integration, time alignment and evidence handling require individual validation.
Connect video analysis with organizational response workflows
Bring video, audio, text, review clips, staff comments and response history together by incident to support internal review and reporting. AI assists factual organization; people retain final judgment and approval.
Images are configuration concepts, not finished product specifications, deployment records or detection-accuracy evidence. AI supports human and organizational review and judgment; it does not determine legal assessments.
SUPPORTED DEVICES
Device options
Commissioned-development scope
Image text translation
Reference translation prepared for this website. The source image is unchanged; this is not an original or certified translation.
1. Using existing surveillance cameras
Analyze installed-camera footage with AI to support incident discovery, clipping, objective review and evidence organization.
- Import existing surveillance footage / source counter sign: Information
AI analysis and detection with automatic clipping of relevant footage
- Video analysis → AI detection → evidence organization
- Dashboard / detections / video search / clip collection / reports / settings
- Detection list: date range 2024/05/15–2024/05/15; all cameras; scene columns: thumbnail, detection type, time and duration
- Abusive language / intimidation, 14:32:18, 00:00:45; prolonged complaint, 14:36:02, 00:01:02; throwing objects, 14:40:29, 00:00:31; contact with an employee, 14:42:55, 00:00:27
- Selected scene: 00:00:06 / 00:00:45. Scene data: 2024/05/15 14:32:18; store camera 1; service counter; abusive language / intimidation; confidence 92% (high)
- Export clip / add to report
Organize and preserve evidence for objective review
- Visualize facts: use footage and data for objective fact checking
- Central evidence management: gather footage, information and reports by case
- Tamper prevention and preservation: access control and logging to support secure retention
- Automatic AI detection of scenes that may involve harassment
- Clip relevant scenes to suitable lengths, automatically including surrounding context
- Organize records by incident for report output and sharing with the relevant parties
- Support early response, consideration of recurrence-prevention measures and operational improvement
The image flows from field footage on the left through the detection interface in the center to human review and preservation on the right. Four supporting functions are listed below.
Site note: 92% is a conceptual-screen value, not measured detection accuracy. The example does not guarantee legal evidentiary value, immunity from tampering or compatibility with every camera.
Existing surveillance cameras
Video analytics using existing equipment
Image text translation
Reference translation prepared for this website. The source image is unchanged; this is not an original or certified translation.
1. Harassment detection and recording device: patent-use platform
AI detection and recording using audio and video: protecting workers and supporting organizational responses.
- On-site devices (collaboration partners): wearable cameras, badge cameras, body cameras, smartphone apps, IP cameras and others
- Standard Link platform: harassment detection and recording device (Japanese Patent No. 7728631); audio and video capture, AI detection, automatic retention from before detection until the incident ends; cloud / secure data infrastructure
- Application settings (adopting organizations): retail and food service, healthcare and care services, transport, finance and service counters, education and local government, other service industries
- Collaboration partners: device manufacturers, software companies, telecom operators, system integrators and security companies
A platform linking technology, data and an ecosystem to co-create real-world deployment.
2. A harassment-response platform expanded through co-creation
Working with diverse partners across field environments.
- Device partners: wearable cameras, badge cameras, body cameras, smartphones, security cameras and others
- Center: Standard Link platform / harassment detection and recording device (Japanese Patent No. 7728631) / detection, recording and retention / secure data infrastructure
- Solution partners: SaaS providers, system developers, communications / network companies, security companies and consulting firms
- Deployment / application fields: retail and food service, healthcare and care services, transport, finance and insurance, education and local government, service industries generally
Combining partners' strengths with our patent technology to co-create workplaces where people can work with confidence.
3. An open platform centered on patent technology
Share detection and recording mechanisms for application in partners' products and services.
- Device / solution collaboration examples: wearable cameras, badge devices, applications, IP cameras and existing-system integration
- Standard Link platform: harassment detection and recording device (Japanese Patent No. 7728631), audio / video capture, AI detection engine, automatic recording / retention control, data management / security, API / SDK provision
- Implementation in partner services: monitoring, cloud services, business-app integration, reports / analysis, insurance / compensation services and others
- Joint development / technical collaboration / marketing collaboration / sales-channel collaboration
Bringing the patent-protected core technology into real-world use through an ecosystem.
4. A platform connecting field sites and supporting peace of mind
Integrating diverse devices and services.
- Field devices: wearable cameras, badge cameras, body cameras, smartphones, cameras and sensors
- Standard Link platform / harassment detection and recording device (Japanese Patent No. 7728631) / AI detection, automatic retention and security management
- Connected partner services: cloud storage, video management, notifications / alerts, analysis / reports and business-system integration
- Application settings: retail, healthcare / care services, transport, education, local government / public services and others
5. Harassment-response infrastructure evolving with partners
Using patent technology as a foundation to enhance product and service value.
- Partner examples: device manufacturers, software companies, telecom operators, security companies and service providers
- Standard Link platform / harassment detection and recording device (Japanese Patent No. 7728631) / audio and video capture, AI detection, retention from before detection, secure management
- Intended value: workplace confidence and safety, faster fact checking, appropriate response decisions, lower risk and increased trust
- Deployment fields: retail / food service, healthcare / care services, transport, finance / service counters, education / local government and service industries generally
An openly connected platform for real-world deployment.
Panel 1 runs devices → core platform → applications, supported by partners below. Panel 2 links both partner groups bidirectionally with the center. Panel 3 connects inputs → core → partner services. Panel 4 flows downward from devices through the platform and connected services to applications. Panel 5 links partners → core → value, with deployment sectors below.
Site note: the five panels are development / collaboration concepts. They do not establish that every device, service or function falls within Patent No. 7728631, is already available, or has a contracted partner. The patent scope is described separately using the published claims.
Other cameras and sensors
Selection and integration for site conditions
Fixed cameras and general video analytics are available for discussion as commissioned development separate from this patent.
Standard Linkcan support thesedevelopment themes
AI cameras / video analytics
From using existing cameras to defining and developing new AI-camera and video-analysis functions.
Audio post-processing / transcription / summaries
Process recordings after capture and organize review intervals, transcripts and key points by incident.
Edge AI / integration with Jetson, VIA and other platforms
Integrate on-site inference, closed-network environments and low-latency processing according to operating conditions.
Incident-management dashboard / workflow UI
Design management interfaces for incident lists, review, permissions, notifications and staff collaboration.
Device selection, development and integration
Connect cameras, microphones, portable devices and existing equipment to fit workflows and data requirements.
PoC design / joint development / business partnerships
Define validation scope and evaluation methods, supporting decisions from PoCs through joint development and commercialization.
Examples of inquiries we welcome
AI upgrades for existing products
Add audio/video analysis and event extraction to recording-only equipment.
Discuss product integrationDeveloping new AI cameras and devices
Design cameras, microphones and edge computers for specific site conditions.
Discuss camera developmentAI management platforms
Manage incidents, video, notifications and permissions across sites in one interface.
Discuss platform developmentIndustry PoCs and joint businesses
Validate in airports, healthcare, hospitality or public counters, then assess sales and commercialization.
Discuss co-creationFrom field inputs to organizational response support.
Separate capture, analysis, incident extraction, record preservation and workflow integration, combining only what each site needs.
Potential field applications
Airports and transport information
In-person assistance in high-traffic locations
Dining and hospitality
Fact-finding and information sharing in customer service
Healthcare and care
Reception, information and consultation desks
Public service counters
In-person reception at local-government and public facilities
Service settings
Workflow design for varied in-person services
Security and facility management
Mobile and in-person duties such as patrols and staffed posts
Images are illustrative, not specific products, actual deployments or performance evidence.
Design connections from cameras to AI infrastructure.
AI EYE
CAMERA
Camera and video inputs, including sensing for low-light and site conditions.
Explore the page → EDGE AITECHNOLOGY
On-site image/audio analysis and sensor fusion.
Explore the page → AI SOLUTIONSAI SOLUTIONS
PoCs, commissioned AI development, management interfaces and workflow implementation.
Explore the page → AI INFRASTRUCTUREAI DATA CENTER
Data storage, distributed processing and integration with operational infrastructure.
Explore the page →From consultation to validation and commercialization decisions.
- 01
Discovery
Clarify the site, challenges, existing equipment, data and operating conditions.
- 02
PoC design
Define the scope of cameras/devices, AI processing, storage, notifications and interfaces.
- 03
Validation and evaluation
Assess site suitability and operations, typically over four to eight weeks.
- 04
Commercialization decision
Discuss deployment, product integration, joint development, sales and business partnerships.
Design capture, storage, access and integration separately.
Data capture
Consider continuous or periodic capture and temporary buffering according to the site and applicable law.
Storage and permissions
Define incident intervals, storage locations, access rights, closed-network configurations and activity logs.
Workflow integration
Connect only necessary information to management interfaces, notifications, existing systems and APIs.
Frequently asked questions
Can I ask about fixed cameras on this patent page?
Yes. This page presents fixed cameras and general video analytics as SLL's separate commissioned-development business, not as an explanation of this patent's scope.
Can AI be added to existing cameras and recording devices?
For existing video and audio the organization may lawfully handle, we confirm purpose, access permissions, retention and computing environment before considering incident clipping, timelines and sharing with staff. This is not an individual victim-support service or outsourced legal-compliance service.
Can you develop AI dashboards and smartphone interfaces?
We welcome discussions of management-platform design and development covering incident lists, timelines, video/audio review, permissions and notifications.
Is audio or video recording legally mandatory?
Japan's amended law and guidelines do not impose a blanket audio/video recording requirement. Organizational measures must include consultation channels, fact-finding, employee protection and prevention of recurrence. If records are used, confirm purpose, access rights, retention and personal-data handling. Installing equipment alone does not complete compliance.
Does AI determine harassment or illegality?
People and organizations make the final decisions. AI handles audio, images and other information to support objective preparation of material for review.
How long does a PoC take?
Typically four to eight weeks, designed around scope and site conditions. NDA and closed-network environments can also be accommodated.
Is recording continuous, and how is personal data handled?
This technology is not intended for continuous surveillance of people.
An embodiment in the granted publication describes sensors capturing audio and images continuously or periodically, deleting or overwriting buffered data after a set time, and preserving at least the interval from before detection through the end of the conduct when harassment is detected.
Actual operations are designed to comply with Japan's Act on the Protection of Personal Information and other applicable laws.
Is the technology only for customer harassment?
No. The specification describes various forms of harassment in workplaces, schools and other settings, including sexual harassment, power harassment, gender harassment and customer harassment.
Devices, cameras, AI and platforms.
Tell us what you need to discuss.
Describe your current challenge, existing equipment and data, and the intended setting. A finished specification is not required at the outset.
Start an inquiry
Describe your current challenge, existing equipment and data, and the intended setting. A finished specification is not required at the outset.
Contact us