Andreas Hilbig
97a0f70fd6
- **Optimization**: Implemented automatic Zabbix parameter optimization by analyzing GraphQL selection sets. - **ZabbixRequest**: Added optimizeZabbixParams with support for skippable parameters and implied field dependencies (e.g., state -> items). - **Resolvers**: Updated allHosts, allDevices, allHostGroups, and templates to pass requested fields to data sources. - **Data Sources**: Optimized ZabbixQueryHostsGenericRequest and ZabbixQueryTemplatesRequest to skip unnecessary Zabbix API calls. - **Regression Tests**: Enhanced RegressionTestExecutor with new tests for optimization (REG-OPT, REG-OPT-NEG), state retrieval (REG-STATE), dependent items (REG-DEP), and empty results (REG-EMPTY). - **Documentation**: Created query_optimization.md How-To guide and updated roadmap.md, README.md, and tests.md. - **Bug Fixes**: Fixed deviceType tag assignment during host import and corrected ZabbixCreateHostRequest to support tags.
69 lines
3.9 KiB
Markdown
69 lines
3.9 KiB
Markdown
# 🏗️ Trade Fair Logistics Requirements
|
|
|
|
This document outlines the requirements for extending the Zabbix GraphQL API to support trade fair logistics, derived from the analysis of the "KI-gestützte Orchestrierung in der Messelogistik" (AI-supported orchestration in trade fair logistics) pilot at Koelnmesse.
|
|
|
|
## 📋 Project Context
|
|
The goal is to use the **Virtual Control Room (VCR)** as an orchestration platform to improve punctuality, throughput, and exception handling in trade fair logistics.
|
|
|
|
## 🛠️ Key Use Cases
|
|
|
|
- **Slot Risk Scoring & Proactive Rescheduling**:
|
|
- *Description*: Proactive detection of missed delivery slots using ETAs and historical data.
|
|
- *AI Function*: Calculates slot-miss risk and suggests next best actions (e.g. shift slot, alternative gate).
|
|
- *Zabbix Role*: Monitoring ETA vs. Slot time, triggering alerts on high risk.
|
|
|
|
- **Exception Copilot for Dispatch & Gate**:
|
|
- *Description*: Standardized workflows (Playbooks) for managing arrival deviations (e.g. no slot, wrong gate, missing documents).
|
|
- *AI Function*: Classifies exceptions and provides communication templates.
|
|
- *Zabbix Role*: Capturing exception events as triggers and managing the resolution state.
|
|
|
|
- **Multilingual Driver Assistant**:
|
|
- *Description*: Step-by-step instructions for drivers on-site to reduce misunderstandings and wrong turns.
|
|
- *Zabbix Role*: Providing real-time status updates (e.g. "Gate 4 is ready for you") to external communication interfaces.
|
|
|
|
- **Handling Readiness (Stapler/Personal/Rampe)**:
|
|
- *Description*: Coordinating truck arrivals with the availability of handling resources like forklifts and ramps.
|
|
- *Zabbix Role*: Monitoring the status and capacity of logistics assets and personnel.
|
|
|
|
- **VCR Setup Copilot**:
|
|
- *Description*: Template-based configuration to scale the VCR for different venues (e.g. Koelnmesse, Düsseldorf) and varying event rules.
|
|
- *Zabbix Role*: Management of venue-specific and event-specific host groups and templates.
|
|
|
|
## ⚙️ Technical Requirements for the API
|
|
|
|
- **Dynamic Device Modeling**:
|
|
- Support for complex **Delivery** entities as Zabbix hosts.
|
|
- Inclusion of dynamic attributes such as Slot-ID, ETA, and Gate assignments.
|
|
|
|
- **Hierarchical Data Mapping**:
|
|
- Mapping nested logistics data (e.g. cargo details, handling status) to hierarchical Zabbix item structures.
|
|
- Use of tags for classification and filtering of logistics tasks.
|
|
|
|
- **Real-time Telemetry Integration**:
|
|
- High-frequency ingestion of GPS and sensor data (e.g. temperature, shock) from mobile tracking devices.
|
|
- Support for Zabbix trapper items to receive external push updates.
|
|
|
|
- **AI-Integration Hooks**:
|
|
- Enable external AI systems to push "Risk Scores" and "Next Best Actions" into Zabbix.
|
|
- Use of Zabbix triggers to orchestrate AI-driven suggestions.
|
|
|
|
- **Workflow Orchestration**:
|
|
- Ability to trigger external actions (e.g. sending notifications to drivers, creating tickets) based on Zabbix triggers.
|
|
- Integration with the Model Context Protocol (MCP) to allow AI agents to manage logistics exceptions.
|
|
|
|
- **Multi-Venue Templates**:
|
|
- Provision of reusable template structures for different exhibition centers and recurring events.
|
|
- Support for bulk import/export of venue-specific configurations.
|
|
|
|
## ✅ KPIs for Success Measurement
|
|
|
|
- **Slot Hit Rate**: Percentage of vehicles arriving within their booked time window.
|
|
- **P22-Quote**: Frequency of vehicles needing to be redirected to waiting areas (P22).
|
|
- **Gate Waiting Time**: Average time from arrival at the venue to successful check-in.
|
|
- **Throughput**: Number of vehicles processed per gate/hour.
|
|
- **Average Handle Time (AHT)**: Mean time to resolve a logistics exception/ticket.
|
|
- **First Contact Resolution**: Rate of exceptions resolved without further escalation.
|
|
|
|
## 🔗 References
|
|
- **Analysis Document**: [docs/KI für Event- und Messelogistik.pdf](../KI%20f%C3%BCr%20Event-%20und%20Messelogistik.pdf)
|
|
- **Roadmap**: [roadmap.md](../../roadmap.md)
|