Cloud Peng

Telemetry Pipeline

Designed equipment telemetry flows that connect industrial devices to backend services, persistent storage, dashboards, reports, and operational audit trails.

Equipment / Edge Service
  -> MQTT Topics
  -> Ingestion Service
  -> MySQL / Redis
  -> Dashboards / Reports / Audit Logs

• Flow: Equipment / edge service -> MQTT topics -> ingestion service -> MySQL / Redis -> dashboards and reports.
• Reliability concerns: heartbeat tracking, configuration synchronization, retry-aware ingestion, offline recovery, and event history.
• Scale path: message rate, machine count, data retention, partitioning, cache strategy, and database indexing can be tuned independently.

RBAC and Governance Model

Built permission governance for internal industrial platforms where traceability, role separation, and operational auditability matter.

User / Department
  -> Role Assignment
  -> Permission Inheritance
  -> Effective Permission Cache
  -> API Authorization / Audit Log

• Model: users, departments, roles, permissions, inheritance, effective-permission cache, and operation audit logs.
• Failure handling: cycle prevention, cache validation, API behavior tests, and database-load testing for permission checks.
• Business value: enables controlled delegation across operations, engineering, IT, and management users.

Industrial Deployment Topology

Delivered systems that can be deployed, operated, and maintained in factory and internal infrastructure environments instead of only local development setups.

Request path: Nginx -> Gunicorn / Flask -> MySQL / Redis
Service management: Linux / systemd
Delivery: Docker Compose / Offline Packages / Runbooks

• Topology: Request path: Nginx -> Gunicorn / Flask -> MySQL / Redis. Linux/systemd manages service lifecycles.
• Operations: Docker Compose packaging, Linux service runbooks, certificate handling, offline packages, and release scripts.
• Infrastructure: Fortinet networking, VLAN planning, access control, Microsoft 365 administration, and collaboration services.

Remote Equipment Operations

Integrated remote access and support tools into engineering workflows for equipment operation, diagnostics, file handling, and field support.

Engineer / Operator
  -> WebSSH / noVNC / File Tools
  -> Equipment Service Layer
  -> Runtime State / Logs / Upload Pipelines
  -> Support Workflow

• Tools: WebSSH, noVNC, Socket.IO, file management, batch operations, and service monitoring.
• Use cases: tester operation, simulator services, remote troubleshooting, engineering data upload, and runtime status inspection.
• System thinking: combines backend workflow design, equipment state modeling, deployment packaging, and operational access control.

Industrial Equipment Data and Utilization Platform

Led architecture and independently developed the platform, now in production use at customer sites, with ownership of requirements, frontend/backend integration, deployment, and maintenance.

• Integrated equipment ingestion, status and utilization monitoring, alarms, reporting, RBAC, and audit trails into a consistent operations interface.
• Designed data persistence, retry, and duplicate handling for recovery from connection or service interruptions.
• Built update checks, service-health verification, failure recovery, and corresponding automated test scenarios.
• Recent versions use a Vue / TypeScript frontend and Flask APIs with SQLAlchemy, MySQL, Redis, MQTT, and Linux deployment.

Result: Delivered a customer-operated platform spanning equipment information, access governance, and field workflows, with ongoing deployment and maintenance ownership.

Tech stack: Python, Flask, Vue, TypeScript, SQLAlchemy, MySQL, Redis, MQTT, Docker Compose, Linux

WAT Test Software and Equipment Monitoring Platform

Led architecture and independently developed and integrated WAT software, delivered for production use at customer sites, with deployment and maintenance ownership.

• Directly implemented instrument communication, measurement execution, and test-flow control, integrating recipe management, test-site selection, data analysis, and equipment monitoring.
• Integrated customer-supplied test algorithms, with syntax checks and simulator-based validation before dispatch to equipment.
• Built engineering data processing, web operations, remote diagnostics, and offline deployment workflows, unifying test configuration, execution preparation, and maintenance support.

Result: Delivered an integrated WAT workflow from test configuration and pre-dispatch validation to equipment operations and maintenance; test algorithms are supplied by customers.

Tech stack: Python, Flask, SQLAlchemy, Pandas, JavaScript, Linux, systemd

Manufacturing Data Analytics and Diagnostic Reporting Platform

Led architecture and independently developed the analytics and reporting platform, now in production use at customer sites for engineering and operations workflows.

• Integrated measurement and diagnostic data into search, statistical analysis, charting, and export workflows, owning backend APIs and data models.
• Designed precomputation, cache refresh, and recalculation workflows to manage result validity and recurring reporting needs.
• Integrated scheduled jobs, service deployment, and operational documentation from data preparation through analytical presentation.

Result: Delivered a consistent analysis interface and recurring reporting workflow for customer engineering and operations users.

Tech stack: Python, Flask, SQLAlchemy, MySQL, NumPy, Pandas, Docker Compose, Linux

Deputy Manager | Oct 2021 - Present

Raytrex Technologies Co., Ltd.

• Serve as Deputy Manager with small-team people-management and hands-on development responsibilities; led architecture and independently developed core platforms now in production use at customer sites.
• Directly developed WAT instrument communication and measurement execution; integrated customer-supplied test algorithms with syntax checks and simulator-based validation before equipment dispatch.
• Built industrial data ingestion, utilization monitoring, RBAC, and audit trails, addressing data retry, duplicate handling, service interruption, and failed-update recovery with corresponding test scenarios.
• Delivered manufacturing analytics with precomputation, cache refresh, recalculation, scheduled reports, and deployment documentation.
• Independently delivered an engineering data editor and enterprise workflow integration service for production use; own requirements, system integration, deployment, maintenance, and coordination across equipment, operations, and IT.

R&D Engineer | Jan 2015 - Sep 2021

Bovia Co., Ltd.

Joined Bovia at its inception following an invitation from my manager during the ITRI department spin-off. Contributed to early company naming and trademark discussions. Worked part-time during the transition, overlapping with my ITRI role.

• Developed device-side and cloud-side web services using Flask, Laravel, MySQL, JavaScript, jQuery, Bootstrap, and RESTful APIs.
• Built modular device management services for networking, VPN, disk, status, system, firmware, firmware upgrade, device control, streaming, and monitoring workflows.
• Designed RabbitMQ-based communication bridges for heartbeat monitoring, device control, cloud coordination, and long-running device service integration.
• Implemented Linux networking and firewall features using iptables, iproute, pyroute2, netifaces, serial modem control, and VPN service management.
• Developed camera, streaming, ONVIF, and PTZ integrations using Video4Linux, Pelco-D, VISCA, recording services, and watchdog processes.
• Built embedded Linux images and filesystem packages with Yocto, Debian root filesystem customization, systemd services, udev rules, and firmware deployment scripts.
• Implemented hardware/device monitoring services for temperature, battery, attitude sensor data, LED/status display, shared-memory status dumps, and health checks.
• Containerized and deployed services across heterogeneous servers and embedded devices using Docker, Gunicorn, systemd, and release packaging scripts.
• Managed Jira, Confluence, and Bitbucket on Google Cloud for engineering collaboration and cloud-side engineering operations.
• Improved system performance through query optimization, load balancing, runtime tuning, and production troubleshooting.
• Mentored junior engineers and supported technical decision-making across backend, embedded Linux, and device-side projects.
• Provided post-employment system maintenance and technical support under a nearly two-year maintenance contract.

Software Developer | Apr 2015 - Jul 2016

Industrial Technology Research Institute

• Developed Flask-based frontend and backend services for device-site applications.
• Built BeagleBone Black images using Yocto for embedded system deployment.

Research and Teaching Assistant | 2012 - 2016

Pervasive Computing Lab

• Used OpenStack to provision big data compute clusters for research and teaching environments.
• Worked with Hadoop MapReduce and distributed file system workflows.
• Used Mesos for two-level resource management in distributed computing environments.

Earlier Experience

• Software Developer (part-time 2012), 新敏科技研發股份有限公司
• Software Developer (part-time Jul 2009 - Jun 2011), Motech Industries Inc.