Phloem
Phloem is a hardware-testing platform that connects bench equipment—DAQ units, SCPI instruments, PLCs, and CAN buses—automatically, with no configuration files or driver hunting. It provides a unified timeline for live monitoring of telemetry, video, and point clouds, and enables deterministic test sequences with local safety loops that run even offline. AI agents read test history, diagnose patterns, and propose next experiments with cited evidence awaiting user sign-off.
- Industrial Automation
- Manufacturing / Industry 4.0
- AI Agents
- Artificial Intelligence
- Aerospace
- Software Only
- Internet of Things
- Data & Analytics
- Hardware
- Robotics
Funding
Founders
Founder details are not available yet.
Product
Problem
Hardware engineering teams face fragmented tooling for test execution, data acquisition, and analysis, forcing them to manually configure drivers, align time-stamped data from disparate sources, and recreate test conditions to debug issues. This slows iteration cycles and makes it difficult to understand root causes when tests deviate from expectations.
Solution
Phloem is a hardware-testing platform that connects any bench equipment—DAQ units, SCPI instruments, PLCs, and CAN buses—automatically, with zero configuration files or driver hunting. It provides a unified timeline for live monitoring of telemetry, video, and point clouds, and enables deterministic test sequences with local safety loops and abort conditions that run even offline. After a run, engineers can replay synchronized data, compare against historical results, and use an AI agent to ask plain-language questions that return answers backed by evidence from test history. Phloem deploys entirely within a company's infrastructure, keeping data on-site and supporting offline-first operation. The platform also includes a context graph that links code, CAD, BOM, requirements, tickets, and test runs, allowing agents to surface impact analysis and draft retest plans automatically.
Target Audience
Phloem is built for hardware engineering teams in industries like aerospace, robotics, and industrial automation that conduct complex test campaigns and need repeatable, data-rich results for rapid iteration.
Features
- Automatic device discovery for DAQ systems, SCPI instruments, LabJack, Ethernet/IP PLCs, and CAN networks with named, typed channels and units
- Real-time visualization of telemetry, video, and point clouds on one synchronized timeline without manual clock alignment
- Deterministic test automation engine with local safety loops (250 ms response), abort conditions, and offline execution capabilities
- AI agents that read test history, diagnose patterns across runs, simulate fixes, and propose next experiments with cited evidence awaiting user sign-off
- Context graph linking SolidWorks, Teamcenter, Confluence, Ansys, Simulink, GitHub, NI DAQ, Jira, Slack, and Phloem runs to trace impact and flag retest requirements
- Synchronized run replay with event logging and comparison against historical runs to identify configuration or signal changes
- On-premise deployment with optional cloud sync, requiring no internet connection for core operations
- A "Trellis" engine for writing test stand logic in code, with a safety loop that trips outputs to safe state in milliseconds