US Software Professionals Inc. • Since 2003
Capability Brief
IDJJ & IDOC — Edge AI
On-premises computer vision for real-time youth safety monitoring, staff ratio compliance, and incident detection — without facial recognition, without the cloud, fully BIPA compliant.
Prepared for the Illinois Department of Juvenile Justice (IDJJ) & Illinois Department of Corrections (IDOC)
Illinois correctional and juvenile justice facilities face mounting pressure to improve safety outcomes while respecting the civil liberties of youth and incarcerated individuals. Senate Bill 1366 and the Illinois Biometric Information Privacy Act (BIPA) create specific requirements around youth safety monitoring and biometric data protection.
SB1366 Context: Illinois legislation emphasizing youth safety, staff accountability, and enhanced monitoring in juvenile justice facilities. VionOS.ai directly addresses these requirements through automated, privacy-compliant safety monitoring that provides continuous oversight without relying on manual processes or biometric surveillance.
Traditional approaches — manual headcounts, human-only CCTV monitoring, paper-based logs — are labor-intensive, error-prone, and provide no real-time alerting capability. Facial recognition systems create BIPA liability and civil liberties concerns that make them unsuitable for Illinois facilities. A modern, privacy-compliant alternative is needed.
VionOS.ai is USSP's on-premises Edge AI platform that transforms existing camera infrastructure into an intelligent safety monitoring layer. It provides real-time person detection, zone occupancy analytics, and automated alerts — all without facial recognition, without biometrics, and without sending video to the cloud.
Key Principle: VionOS.ai uses skeletal pose estimation and bounding-box tracking — never facial recognition. Each detected person is assigned an Anonymous Person Token (SHA-256 hashed, 4-hour TTL) for spatial tracking without identification. Zero biometric data is ever collected, stored, or transmitted.
The platform runs entirely on NVIDIA Jetson Orin edge devices installed at the facility. Video frames are processed locally in real time and never leave the facility network. The system generates structured event data (anonymous counts, zone occupancy, alert triggers) that feeds a real-time dashboard accessible to facility administrators and oversight personnel.
Real-time counting of staff and youth in each zone. Automated alerts when ratios fall below required thresholds (e.g., 1:8, 1:12 depending on zone classification). Continuous compliance logging with timestamps for audit and oversight reporting.
Instant alerts when a zone containing youth has no staff presence for a configurable time threshold. Prevents unsupervised situations before they escalate. Configurable per zone, per time period, per facility policy.
Immediate alerts when any person enters a restricted or off-limits zone. Configurable per zone, per time of day, per access level. Supports secure perimeter monitoring and contraband-prone area oversight.
Detects clustering (multiple people converging rapidly), rapid dispersal patterns, and unusual dwell time that may indicate developing incidents. Early warning system without identification — alerts staff to investigate before situations escalate.
Continuous anonymous headcounts by zone, updated in real time. Entry/exit logs for movement between areas. Eliminates manual count errors and provides a continuous record of population distribution across the facility.
Automated monitoring during high-risk shift change periods. Ensures zone coverage continuity during staff handoffs. Verifies that incoming shift achieves required staffing ratios before outgoing staff depart zones.
VionOS.ai was designed from the ground up to comply with the Illinois Biometric Information Privacy Act (BIPA, 740 ILCS 14). The system architecture ensures zero biometric data collection at every layer of the technology stack.
| Privacy Requirement | VionOS.ai Implementation |
|---|---|
| No facial recognition | Skeletal pose estimation + bounding-box tracking only. No face detection, encoding, or matching algorithms. |
| No biometric data collection | Anonymous Person Tokens (SHA-256 hash, 4-hour TTL). No fingerprints, voiceprints, iris scans, or facial geometry. |
| No PII storage | No names, faces, voices, or identifiers. Only anonymous spatial and behavioral event data. |
| No cloud transmission | All processing on local NVIDIA Jetson hardware within the facility network perimeter. |
| No video recording/egress | Raw video frames processed in real time and discarded. Never stored, never transmitted outside facility. |
| Audit trail | Timestamped event logs with zone IDs and anonymous tokens. Configurable retention policies. |
| Data sovereignty | All data remains on facility-controlled hardware. No third-party data processors. |
BIPA Safe Harbor: Because VionOS.ai never collects, captures, or stores biometric identifiers or biometric information as defined in 740 ILCS 14/10, the system operates outside the scope of BIPA's consent, disclosure, and data retention requirements. No written consent from monitored individuals is required.
Web-based dashboard on the facility network showing live zone occupancy, staff ratios, and active alerts. Role-based access for administrators, supervisors, and correctional officers. Shift handoff reports included.
Configurable alert thresholds per zone — staff ratio violations, unattended zones, restricted area intrusions. Alerts delivered via on-screen notification, email, and SMS.
Timestamped event logs and audit trails for oversight reviews. Historical trend analysis for staffing patterns, incident frequency, and zone utilization. Exportable reports.
No new camera infrastructure required. VionOS.ai works with existing IP cameras over the facility network. Compact edge devices install in a server closet or rack — no data center needed.
Issue via TOPS Category Two (CMT4599470)
Facility, camera, and requirements evaluation
On-premises pilot, 1-2 priority use cases
Performance tuning, stakeholder sign-off
Phased rollout to additional facilities
Procurement: IDJJ and IDOC can engage USSP through the TOPS Category Two contract (CMT4599470, BidBuy P-59947). USSP is an authorized subvendor to Krasan Consulting Services, the Primary Awardee. Agencies issue a Task Order through the existing TOPS vehicle — no separate RFP is needed. The contract is available through February 2034.
A typical pilot deployment covers 1-2 priority use cases (e.g., staff-to-youth ratio monitoring + unattended zone detection) in a single housing unit or dayroom. USSP provides temporary edge hardware for the pilot at no hardware cost to the agency. The pilot validates detection accuracy, alert reliability, and operational fit before the agency commits to full procurement and facility-wide deployment.
After pilot validation, VionOS.ai scales incrementally — additional zones within the pilot facility, then to additional IDJJ/IDOC facilities. Each expansion follows the same assess-deploy-validate cycle. Centralized multi-facility dashboards are available for statewide oversight.
Established January 23, 2003 in Chicago, IL. Two decades delivering technology solutions to government agencies, healthcare organizations, and Fortune 500 enterprises.
Illinois BEP-certified Minority Business Enterprise (MBE) and Women Business Enterprise (WBE). IDOT-certified Disadvantaged Business Enterprise (DBE). Helps agencies meet supplier diversity goals.
Authorized subvendor under CMT4599470 to Krasan Consulting Services. TOPS covers professional services — assessment, configuration, deployment, training, and support staffing. USSP provides temporary hardware for the pilot phase. For scaled deployment, the agency procures edge hardware through its standard procurement process. Available through February 2034.
Assessment, pilot, deployment, training, and ongoing support. A single partner from initial facility evaluation to scaled statewide operations across multiple IDJJ/IDOC facilities.