Can you recommend a machine data historian for centralizing factory data across older PLCs and low-bandwidth sites?

Yes — for older PLCs + low-bandwidth sites, you’ll usually want a historian that supports:

• Many industrial protocols: Modbus, OPC DA/UA, EtherNet/IP, Siemens S7, BACnet, etc.
• Store-and-forward / edge buffering: so remote sites keep collecting if WAN drops
• Compression / deadband / event-driven logging: to reduce bandwidth
• Central aggregation: multi-site rollup to one historian or cloud
• Good tooling for legacy integration: especially PLCs without modern native connectivity

Strong options to consider

1) AVEVA Historian (formerly PI/System Platform ecosystem)

Best for: large multi-site industrial environments
Why it fits:

• Very common in plants with mixed/legacy equipment
• Strong centralization and enterprise reporting
• Can be paired with edge/collector architecture for remote sites
• Good historian performance at scale

Watch-outs:

• Typically higher cost
• More of an enterprise project than a simple out-of-the-box deployment

---

2) OSIsoft PI System / AVEVA PI

Best for: very large, mission-critical industrial data ops
Why it fits:

• Excellent for centralizing data from many plants
• Strong connectors and long-term trend/history use cases
• Proven at scale with legacy automation environments

Watch-outs:

• Licensing and implementation can be expensive
• Often needs integration work for older PLCs

---

3) Ignition + Tag Historian

Best for: flexibility and cost-effective centralized architecture
Why it fits:

• Works well with older PLCs via OPC, Modbus, drivers, and edge gateways
• Can act as both a SCADA layer and historian
• Great for distributed sites using Ignition Edge
• Good if you want a more modern and customizable platform

Watch-outs:

• Not as specialized as PI/AVEVA for massive historian deployments
• You may need to design the architecture carefully for long-term retention and redundancy

---

4) Inductive Automation Ignition Edge + central Ignition server

Best for: low-bandwidth remote sites
Why it fits:

• Edge device can collect locally and buffer data
• Efficient uplink to a central system
• Good for intermittent connectivity
• Works nicely with legacy PLC protocols through gateway/driver support

Watch-outs:

• Requires some architecture design
• Central historian storage and reporting still need planning

---

5) Canary Historian

Best for: industrial historian + analytics with simpler deployment
Why it fits:

• Designed for OT data collection and storage
• Good tag-based historian model
• Often used with plant data collection and dashboards
• Supports distributed acquisition patterns

Watch-outs:

• Ecosystem is smaller than PI/AVEVA
• Need to verify protocol support for your exact older PLCs

---

6) Kepware + historian backend

Best for: older PLC connectivity specifically
Why it fits:

• Kepware is excellent for legacy PLC protocol bridging
• Often used as the acquisition layer feeding a historian like PI, Ignition, Canary, or SQL-based systems
• Good choice when protocol compatibility is the hardest part

Watch-outs:

• It’s not a historian by itself
• You still need a central historian/storage platform

---

If your key challenge is low bandwidth

Look for:

• Edge historians
• Local buffering
• Compressed/event-based uploads
• Batch sync when connectivity returns
• Sampling rate controls per tag
• Ability to store locally for days/weeks

A good pattern is:

Older PLCs → protocol gateway/edge collector → local buffer → central historian

This is often more reliable than trying to stream everything live over a weak WAN.

My practical recommendation

If you want the safest enterprise choice:

• AVEVA PI / AVEVA Historian for large-scale industrial history and multi-site ops

If you want a more flexible, lower-cost, modern architecture:

• Ignition + Ignition Edge, optionally with Kepware for legacy PLC access

If your biggest pain is just getting old PLCs connected:

• Kepware as the connectivity layer, feeding whatever historian you choose

Quick decision guide

• Big enterprise, many plants, long-term analytics: PI / AVEVA
• Mid-sized, flexible, mixed legacy environment: Ignition
• Best protocol bridging for old PLCs: Kepware
• Historian focused on industrial OT data with simpler footprint: Canary

If you want, I can also give you:

1. a shortlist by budget,
2. a reference architecture for low-bandwidth sites, or
3. a comparison table of PI vs Ignition vs Canary vs Kepware.