NATIVE INTEGRATIONS
Haltless integrates seamlessly with the world's leading industrial hardware and enterprise software. Explore our native compatibility below.
01 · CONNECTIVITY
Fifteen ways to read your machines
Fourteen input protocols and one direct push API. The Python edge agent that ships with the platform speaks all of them.
OPC UA
Any compliant OPC UA server, including the latest spec with security policies and message signing.
Modbus TCP
Any Modbus TCP slave; supports holding registers, input registers, coils, and discrete inputs.
Modbus RTU
Modbus devices on an RS-485 or RS-232 serial line, usually through a USB adapter. Same registers and decoding as Modbus TCP, each device on the bus by its unit ID.
Siemens S7
S7-300 and S7-400 read directly over S7comm, S7-1200 and S7-1500 with PUT/GET access enabled. No OPC UA server or gateway in between.
EtherNet/IP
Rockwell and Allen-Bradley over CIP: ControlLogix, CompactLogix and Micro800 by tag name, SLC 500 and MicroLogix by data-file address.
Mitsubishi MELSEC
Q, L, iQ-R and iQ-F (FX5) CPUs over the MC protocol (SLMP), word and bit devices addressed as the CPU numbers them. Batch reads only: nothing is written to the PLC.
Omron FINS
CJ, CS and CP CPUs, and NJ/NX variables in CJ-compatible memory, over FINS/UDP. DM, CIO, Work, Holding and EM addresses as Omron writes them, read with Memory Area Read only.
Beckhoff ADS
TwinCAT 2 and 3 PLC variables by symbol name, all read in one sum command per cycle, with a reconnect when a download moves them. The agent never adds a route on the PLC.
BACnet/IP
Building controllers such as air handlers, chillers and boilers: present values read with ReadPropertyMultiple, binary and multi-state objects as numbers. Nothing is written.
MQTT
Subscribe to topics on the broker your plant already publishes to, plain values or JSON, with a login, TLS or a client certificate. Nothing is published back.
Sparkplug B
The edge nodes and devices of a Sparkplug B group on your broker, decoded from births, data and deaths, with units from the birth. No rebirth requests and no commands.
AVEVA PI Web API
Current values of PI points and AF attributes over HTTPS, with PI's own timestamps and units. GET requests only, so the historian stays the system of record.
CSV file ingestion
Watch a directory; pick up MES exports and historian CSV dumps automatically after every shift.
JSON file ingestion
Same file-watch flow for JSON drops from legacy production systems and homegrown data pipelines.
REST API
Direct push from custom integrations using scoped, read- and write-capable API keys.
02 · PLC AND HMI COMPATIBILITY
Any PLC that speaks OPC UA, Modbus, S7, EtherNet/IP, MELSEC, FINS or ADS
Siemens, Rockwell, Mitsubishi, Omron and Beckhoff PLCs are read natively, and any controller that exposes OPC UA or Modbus works too, so the list below is illustrative rather than exhaustive. If your PLC speaks one of these protocols, the edge agent reads it.
03 · ENTERPRISE SYSTEMS
Connect to the systems your business already runs
ERP, identity, and the historians your engineers already trust.
ERP adapters
Production adapters for SAP S/4HANA, Oracle EBS and Microsoft Dynamics 365: bi-directional work orders, parts reservations and cost rollups, with the same audit chain behind every write.
Identity and SSO
Tenant-level OpenID Connect with PKCE for any OIDC identity provider, including Okta, Auth0, JumpCloud, Microsoft Entra ID and Google Workspace.
Historian bridges
OSIsoft PI and AVEVA PI reachable through their OPC UA bridges. Inductive Automation Ignition through its built-in OPC UA server.
04 · NOTIFICATIONS AND WORKFLOW
Ten outbound channels to the rest of your stack
Every alert routes to wherever your team already lives. We support these channels out of the box.
SMTP-based transactional email with templated bodies, retry on bounce, and per-tenant signing.
SMS
Short-message escalation through a regional SMS gateway, throttled per machine to keep noise out of the on-call.
Webhook
Signed POST to any HTTPS endpoint, with retries and a published payload schema.
05 · BI AND REPORTING
Plug your dashboards into Haltless data
A read-only REST API plus CSV exports for every report keep your BI stack in the loop.
REST API with scoped keys
Read-only API keys with fine scopes such as read:machines and read:sensors, so a BI integration cannot reach beyond what it needs.
CSV exports per report
Fleet, OEE, maintenance, downtime, and work-orders reports all export as CSV from the dashboard or the API.
Grafana panels
Pull machine health and OEE figures through the REST API into Grafana for your existing operations dashboards.
Power BI and Tableau
Connect to the REST API and refresh your Power BI or Tableau dashboards on the cadence your analytics team wants.
COEXISTENCE
Haltless does not replace your historian, SCADA, or CMMS
Haltless is designed to live next to the systems your plant already trusts, not on top of them. Three concrete patterns.
Historians stay your system of record
If OSIsoft PI or AVEVA PI is your long-term tag store, Haltless reads current values straight from the AVEVA PI Web API, or through an OPC UA bridge. We do not replicate years of historical data; we score the live readings, and nothing is written back to the historian.
SCADA and DCS stay in control
Haltless is a monitoring and analytics platform, not a control system. We do not write to PLCs, we do not close control loops, and we do not coordinate emergency-stop arrangements. Your existing SCADA or DCS keeps that responsibility.
CMMS hand-off, not displacement
If you already run MaintainX, UpKeep, Fiix, IFS, or Maximo, Haltless routes alerts in as work orders through the ten outbound notification channels (ServiceNow and Jira included) rather than forcing them into our own CMMS. The choice is per tenant.
COMMON QUESTIONS
Questions about connecting it up
Which industrial protocols does Haltless support?
OPC UA (IEC 62541), Modbus TCP, CSV file watch and JSON file watch are in production, and the edge agent reads Siemens S7 and EtherNet/IP via CIP natively, with no gateway in between. Where data already flows through an MQTT broker, the agent subscribes to its topics. It reads Modbus RTU, Mitsubishi MELSEC, Omron FINS, Beckhoff ADS and BACnet/IP natively as well, decodes Sparkplug B from a broker, and takes current values from the AVEVA PI Web API. If a machine only drops files somewhere, the file watchers are usually the fastest route in.
Will it work with our existing CMMS?
Alerts open work orders in the platform’s own CMMS workflow, and integrate outward to the enterprise systems you already run. Haltless complements SCADA, DCS, MES, ERP and your existing CMMS rather than replacing any of them.
How do alerts reach the people who need them?
Through ten notification channels, so an alert lands where a team already works rather than in one more dashboard nobody opens. The alert lifecycle and every signoff on it are recorded in the audit chain.
How long does it take to get connected?
Install the agent on a Linux box near the machines and point it at the PLCs. There is no training period: the detectors calculate from the first readings they receive. More history makes the baseline sharper, so the same alert rests on more evidence in week three than on day one.
