Line View
Live inspection with the current frame and verdict, a searchable defect gallery, and a reject audit that keeps every image with its model version, confidence and reason.
Perflens Console
Rejects, bearings and kilowatts on one screen
Perflens Console runs on a Gatepoint or a server in your building. Quality, maintenance and facilities teams each get their own view of the same data, and nothing stops working when the internet does.
The operating layer
On-premise inspection and monitoring software, by design
Console is where people meet the Edge Stack. A quality engineer opens it to audit last night's rejects. A maintenance planner opens it to see which gearbox moved from watch to act. An energy manager opens it to find out why the second shift uses more compressed air than the first.
It runs where the data is: on the Gatepoint itself for a single cell, or on an on-premise site server for a whole plant. No module needs an internet connection. Data is stored locally under retention periods you set, and leaves the site only through the exports and integrations you configure.
Every view answers the same three questions in its own terms: what happened, why the system thinks so, and what to do next.
Modules
Six modules, one data model
Each module is a view onto the same asset register, event history and model registry, so a reject spike and a drive fault on the same line appear on the same timeline.
Asset Health
Condition trends per asset, fault diagnostics such as imbalance, misalignment and bearing wear, and a full alert history that records what was found when someone went to look.
Performance
OEE by line and shift, microstop detection and loss trees, including machine states captured from equipment with no PLC. See OEE tracking.
Facility
Energy sub-metering, compressed-air baselines and leaks, HVAC and refrigeration health, environmental conditions and privacy-preserving occupancy counts.
Models
The model registry, a model card for every version, the drift monitor and the retraining queue — what is deployed where, and how well it is performing.
Fleet
Device health, over-the-air update status and certificates for every Lensnode, Pulsenode and Gatepoint, with staged rollouts and one-step rollback.
Alerts
Alerts that reach the right person, with the evidence
An alert flood is how monitoring systems lose their audience. Console routes each alert by asset, severity, shift and role, and every alert carries its own explanation: the signal, the threshold or model output that triggered it, the image or spectrum behind it, and a suggested first check.
Alerts can go to:
- SMS and email for the on-call technician or engineer
- Microsoft Teams and Slack channels for a line or a maintenance crew
- Stack lights on the line, driven directly from Gatepoint outputs
- CMMS work orders through the API of systems such as IBM Maximo, SAP PM, Fiix, Limble or UpKeep, with the evidence attached
Unacknowledged alerts escalate on a timer you set. On Care Verified and Care Dedicated, a Perflens analyst checks the evidence within the tier's verification window — 4 business hours, or 1 hour around the clock — so a confirmed alert means someone has already looked.
Deployment
Three ways to deploy Console
| Mode | Where it runs | Best for | Internet needed |
|---|---|---|---|
| Gateway | On the Gatepoint that runs the models | A single inspection cell or a small group of monitored assets | No |
| Site server | On an on-premise server in your building, physical or virtual | A plant with several lines, buildings or Gatepoints | No |
| Private-cloud aggregation | A Console instance in your own private-cloud tenancy, fed by each site | Multi-site fleets that want one view across plants | Only between sites and the aggregation instance; each site keeps running if the link drops |
Access control
SSO, roles and a full audit log
Single sign-on
SAML 2.0 and OpenID Connect against your identity provider, so joiners and leavers are handled where your IT team already manages them.
Role-based access
Roles scope what each person can see and change, by site, line and module. An operator can acknowledge an alert; only a quality engineer can approve a model change.
Full audit log
Every login, setting change, model deployment, reject override, export and remote-support session is recorded with who, what and when, and can be forwarded to your SIEM over syslog.
Your data
Data export and retention
Your data belongs to you, in formats other tools can read without us.
- CSV
- Any table or filtered view — rejects, alerts, OEE, energy, audit log
- Parquet
- Bulk export of time-series and event data for analytics platforms and data lakes
- Historians
- Live feeds over OPC UA, and MQTT with Sparkplug B
- REST API
- Results, alerts and asset data for your own applications
- Images
- Reject and sample images with their metadata, in standard image formats
- Models
- Every trained model exportable as ONNX, with its model card
- Retention
- Set per data type; storage is sized in your proposal
FAQ
Questions we hear often
Does Console need an internet connection?
No. Every module runs on the Gatepoint or your site server. An outbound connection is only used if you enable remote support or multi-site aggregation, and you can switch either off. See the Security & Trust Center.
Can Console run on our own virtual infrastructure?
Yes. In site-server mode it runs on a physical server or a virtual machine in your data room. We specify the processor, memory and storage it needs in your proposal.
Will Console replace our SCADA or MES?
No. Console is the system of record for inspection decisions, condition data and model history. It feeds your SCADA, MES and historian rather than replacing them.
Can we bring data from other equipment into Console?
Yes. Machine states and process values from PLCs, meters and drives can be read over OPC UA, Modbus, EtherNet/IP or PROFINET and shown beside Perflens sensor data.
How are Console updates delivered?
As signed updates through the Fleet module, scheduled in your maintenance windows, with rollback if anything fails a health check. Our support page sets out the update policy.
Walk through Console with an engineer
Bring your quality, maintenance or facilities lead. In a 45-minute discovery call we'll show the modules that matter to your problem.