Cadence Cockpit • Cognitive Closed-Loop Digital Twin (C-CLDT)

Forenly AI Systems' Entry for the Splunk Agentic Ops Hackathon (2026)

A Physical AI observability platform that maintains a real-time digital twin of an autonomous robot fleet — closing the loop between shop-floor telemetry and autonomous corrective action through Splunk Cloud HEC, GCP Gemini multi-agent orchestration, CAD branching (Onshape), ERP financial ledgers, and on-demand capacity leasing (Deel).

🚀 Live Production Dashboard: https://quality.forenly.ai

---

Community

Building this in the open — join the team chat on Discord: https://discord.gg/7JXRwTy2EE

---

💡 The Core Innovation: Breaking Design-Centricity

Traditional Configuration Management (CM) and Product Lifecycle Management (PLM) systems operate in a design-centric silo. They treat change management purely as an engineering activity—tracking geometric revisions, CAD drawings, and design specifications. This isolated approach ignores two critical real-world axes:

1. The Economic Axis: How global commodity spot-price fluctuations (e.g., London Metal Exchange refined copper) impact active BOM margins and force otonom material substitution to protect profitability.
2. The Organizational Axis: How physical changes on the shop floor trigger otonom machine calibration (G-Code injections) and dynamic global contract leasing (on-demand offshore capacity/manpower via API models).

Cadence Cockpit represents a Cognitive Closed-Loop Digital Twin (C-CLDT). It mathematically couples and automates Engineering (CAD), Economic (BOM Margins), and Organizational (IoT Calibration & Offshore Leasing) channels into a unified, zero-waste cyber-physical loop.

                  ┌─────────────────────────────────────────┐
                  │    Unified Splunk HEC Ingestion Hub    │
                  └────────────────────┬────────────────────┘
                                       │ (Big-Data Telemetry)
                                       ▼
                  ┌─────────────────────────────────────────┐
                  │     Cognitive Gemini IRB Orchestrator   │
                  └────────────────────┬────────────────────┘
                                       │
                    [ Dynamic Multi-Dimensional Triage ]
                                       │
         ┌─────────────────────────────┼─────────────────────────────┐
         ▼                             ▼                             ▼
┌──────────────────┐          ┌──────────────────┐          ┌──────────────────┐
│   Engineering    │          │     Economic     │          │  Organizational  │
│   Change (ECM)   │          │   Change (Eco)   │          │   Change (OCM)   │
├──────────────────┤          ├──────────────────┤          ├──────────────────┤
│ • Onshape CAD    │          │ • Odoo ERP Cost  │          │ • G-Code Offset  │
│ • Geometric      │          │   BOM Ledgers    │          │   Injection (G10)│
│   Tolerances     │          │ • Commodity Spot │          │ • On-Demand Deel │
│ • Branching APIs │          │   Price Index    │          │   Capacity Lease │
└──────────────────┘          └──────────────────┘          └──────────────────┘

---

🛠️ The 3 Simulation Channels

Cadence Cockpit programmatically demonstrates the otonom triage and resolution of manufacturing and financial anomalies across three channels:

1. 📐 Engineering Change Loop (CAD & Design)

• The Anomaly: Solomon 3D Vision edge camera flags a wiring path deviation (NG_WIRING) on PCB board assembly PNL-A12.
• The Triage: The AI Risk Agent evaluates the anomaly and determines that it is a localized, low-risk issue suitable for Fast Track routing (bypassing the CRB).
• The Execution: UiPath RPA calls Onshape SaaS REST APIs to create a design branch Defect-Revision-1, and writes corrective parametric variables ($\Theta_{CAD_New} = \Theta_{CAD} + \Delta\Theta$) back to the CAD model before logging a compliant ECR in the PLM vault.

2. 📈 Economic Change Loop (BOM & Supplier Pricing)

• The Anomaly: Real-time emtia spot trackers flag an Refined Copper price surge on the LME (+35.4%), which degrades the product gross profit margin below the mandatory $20%$ margin floor.
• The Triage: The Gemini IRB Agent flags a margin-floor violation.
• The Execution: Instead of human negotiation, the agent calculates a material substitution option—minimizing copper mass and substituting a thermoplastic vulcanizate (TPV) sleeve—updating the Odoo ERP BOM cost structure automatically.

3. 🌍 Organizational Change Loop (Calibration & Offshore Capacity Leasing)

• The Anomaly: Optical sensors log consecutive dispenser nozzle calibration drifts ($+0.45\text{ mm}$), violating ISO-9001 quality thresholds.
• The Triage: The agent flags an organizational/machinery drift.
• The Execution (The Deel & Dell Model):
  1. IoT Injection: The system dynamically transmits a G-Code offset transformation (G10 L2 P1 X-0.45 Y+0.40) to the machine controller.
  2. LMS Assignment: Automatically schedules a 15-minute calibration retraining course in the operator's LMS.
  3. Offshore Leasing: If local machinery is undergoing heavy maintenance, the system uses on-demand platform APIs (such as Xometry/Deel) to dynamically lease offshore manufacturing capacity in a lower-cost zone, transferring G-Code structures and spinning up temporary local QA auditor contracts otonomously.

---

🔌 Technology Stack & 9-Node Architecture

The platform integrates 9 distributed enterprise nodes into a single, cohesive cyber-physical system:

1. GCP Gemini Orchestrator: The "core mind" that handles CM2 compliance risk equations and orchestrates multi-dimensional triage.
2. Splunk Cloud (HEC): Ingests big-data telemetry from shop floor, market spot tickers, and robot logs.
3. UiPath RPA Bot: Serves as the cross-system execution robot, updating legacy registries, ERPs, and CAD parameters.
4. Qwen-VL Edge AI: Edge visual-language model that scans 3D point clouds and classifies geometric defects.
5. Slack Bolt Socket Mode: Interactive approvals interface allowing Change Specialists to authorize full-track ECRs.
6. Build with Gemini XPRIZE Node: Analyzes total waste reduction, carbon-credits, and energy efficiency.
7. Onshape Cloud CAD: Cloud-native single source of truth for 3D physical geometries.
8. Odoo ERP / Cost Tracker: Real-time financial ledger tracking material pricing and gross margin ratios.
9. CM2 Master Thesis Framework: The underlying scientific and theoretical methodology guiding process audits.

---

🚀 Setup & Run Instructions (Local Sandbox)

The sandbox VM serves the live dashboard on port 8050 and acts as the secure API bridge to Splunk Cloud and enterprise systems.

1. Prerequisites

Ensure you have Python 3.8+ installed on your host. No external pip libraries are strictly required for the lightweight server, as it relies on core Python modules (http.server, urllib, json) to guarantee 100% execution reliability.

2. Configure Environment Variables

Create a .env file in the project root:

PORT=8050
SPLUNK_HEC_URL=http://127.0.0.1:8088/services/collector
SPLUNK_HEC_TOKEN=bf5a8e32-1594-4d92-bb8a-d14c33ea98f1

3. Start the Server

Run the lightweight web-server:

python3 server.py

The server will initialize on port 8050 and print: Forenly Splunk-Agentic-Ops Dashboard serving on port 8050

4. Direct Access & SSH Tunneling

If running on a remote cloud instance, establish a local SSH tunnel to securely interact with the dashboard:

ssh -N -L 8050:127.0.0.1:8050 macb@34.179.172.73

Now, open your local browser and navigate to http://localhost:8050 to interact with the Cockpit.

---

⚖️ License

This project is licensed under the Apache License 2.0 - see the LICENSE file for details.

---

👥 Team

Forenly AI Systems · github.com/forenly-ai-systems