Data-Acquisition System for ISRO’s Cryogenic Engine Test Facility

Back to News & Events

Redundant DAQ System for ISRO’s Cryogenic Engine Tests

In this article, we describe the data-acquisition system developed for the Indian Space Research Organization (ISRO) for their newly designed Integrated Cryogenic Engine & Stage Test (ICET) Facility at the Propulsion Complex of IPRC Mahendragiri in Tirunelveli District, Tamil Nadu, India. While the entire ICET structure was built by TATA Projects, the data-acquisition system was delivered by Bustec. This system comprises approximately 15,000 measurement channels. Its design is based on LXI (LAN eXtensions for Instrumentation) instruments developed by Bustec to meet the specific requirements of IPRC.

The system measures the following parameters: pressure, temperature (RTD and thermocouples), speed, flow, vibration, dynamic and acoustic pressure, and strain.

The physical layout is as follows: all instrumentation cables from the field are terminated in the Cable Terminal Room (CTR), located 60m from the test bay.

Cable Terminal Room with some of the Bustec modules
Cable Terminal Room with some of the Bustec modules

The ICET Control Centre (ICC) is in a secure location 2.5 km away. Communication between CTR and ICC is established via a fibre-optic network.

Main control room at ICC

The actual Cryogenic test facility built by TATA Projects is shown below. 

When Accuracy Matters — Bustec delivers redundancy and reliability.
Actual Cryogenic test facility built by TATA Projects

Every data-acquisition system is fully redundant: each system transmits data via two independent (redundant) Ethernet ports to separate servers. In addition, all instruments provide two analog outputs, which are routed to further servers. Consequently, if any instrument, server, or switch fails, data recording continues uninterrupted.

System requirements

Each test costs several million dollars, so data loss is unacceptable. To mitigate risk, multiple servers collect data.  As noted above, data is recorded both in the CTR and ICC, with several levels of redundancy.

Every data-acquisition system is fully redundant: each system transmits data via two independent (redundant) Ethernet ports to separate servers. In addition, all instruments provide two analog outputs, which are routed to further servers. Consequently, if any instrument, server, or switch fails, data recording continues uninterrupted.

See Figure 1 as an example.

Figure 1. Parallel and redundant data-acquisition systems (Main networks and Redundant subsystem with dual Ethernet paths and analog-output routing).

The system captures data from both the Engine Bay and the Stage Bay, and consists of the following subsystems:

  • High-Speed Data Acquisition System (HDAS)
    • 96 channels of vibration and acoustic pressure measurements with IEPE sensors
    • Sampling rate: 50 kS/s
    • Capable of sustained acquisition, live display, and storage
    • Designed to capture up to 1,200 seconds of data in a single run
  • Unified Data Acquisition System for Facility Monitoring (UDAS-F)
    • 4,608 digital input channels, 96 RTD channels, and 672 voltage measurement channels (various sensors)
    • Sampling rate: 1 kS/s (digital), 100 S/s (RTD), 2 kS/s (voltage)
  • Unified Data Acquisition System for Test Article Monitoring (UDAS-A)
    • 64 frequency measurement channels for flow
    • 224 cryogenic-range RTD channels
    • 112 thermocouple channels
    • 160 strain gauge channels for pressure measurements
    • Sampling rate: 2 kS/s (100 S/s for RTD)

Subsystem Redundancy

  • HDAS: Separate Main and Redundant subsystems, each operated by three computers — server, controller, and additional data-storage station.
  • UDAS-F and UDAS-A: Separate Main1, Main2, and Redundant subsystems, each operated by three computers — server, controller, and additional data-storage station.
  • For UDAS-F and UDAS-A, all data acquired by any subsystem can be merged via software.

The complete system consists of 88 LXI instruments, each supporting up to 256 channels. All instruments are 1U high and 19″ wide, with dual redundant 1 G Ethernet ports and dual redundant power supplies. Most instruments also have two analog outputs per input channel, and all comply with IEEE 1588. Synchronization is maintained to ~10 ns via a GPS Grandmaster clock over the PTP (IEEE 1588) protocol.

As an example, the UDAS architecture is shown in Figure 2. Instruments are numbered 1–55, with three output ports each:

  • Blue and green ports: redundant Ethernet connections (data flow shown in the diagram)
  • Yellow ports: analog outputs routed to additional voltage measurement instruments for recording

At the CTR, data is collected by a primary server, then forwarded over fibre to the ICC server. The ICC also hosts the acquisition controller and viewing stations. If one controller loses connection, control automatically transfers to the backup.

flow chart
Figure 2. UDAS redundancy and data flows between CTR and ICC (instrument ports, redundant Ethernet paths, analog-output routing, controller failover).

ProDAQ System Software

The ProDAQ DAAS application configures, controls, acquires, processes, stores, and graphically displays data from multiple kinds of ProDAQ instruments. All instruments synchronize using IEEE 1588 to tens-of-nanoseconds accuracy. All data are timestamped, ensuring precise correlation even with mixed sampling rates. The software performs live conversion to engineering units and offers FFT and time-domain analysis in both online and offline modes. It can store and reload very large datasets — hundreds of channels and hours of captured data.

The software is optimized for performance and fault tolerance. It provides a detailed live health status of instruments and is not affected by any instrument, network, or client-PC failure. It can also recover parts of the system without affecting ongoing acquisition.

The application can run on multiple computers simultaneously, connecting to the same instruments — possibly on two redundant networks — and can act as an exclusive controller and/or a passive listener to data streams. If the controller side fails, a listener can take over control. An embedded server can provide data and system configuration to other clients, including additional DAAS instances. Other clients can propagate data via specialized interfaces (currently EtherCAT and OPC UA). Clients tailored for special use cases can also be developed.

This flexible approach allows a single software platform, by means of static configuration, to scale into a multi-level redundant system. Through its flexibility and parallelism, the software can be used with a single instrument or with more than one hundred instruments with similar ease of use and performance. There is no practical bandwidth limitation: if throughput requirements grow, additional servers can be added. Functionally, the software behaves like a large real-time database, independent of total data bandwidth or the number of servers.

Summary of Key Features

  1. Additional redundant subsystems measuring analog outputs from all channels, linearized to engineering units (768 and 560 channels, respectively).
  2. Network-level redundancy in UDAS Main1 and Main2, ensuring uninterrupted storage even if one network path or computer system fails (data is identical on both networks).
  3. PSU redundancy for each ProDAQ module.
  4. IEEE 1588 synchronization to a Grandmaster clock for ±10 ns accuracy and UTC traceability, enabling third-party system synchronization.
  5. Unified data display, uninterrupted even if any of the three data sources (two main networks plus redundant subsystem) fails.
  6. Centralized control from a single controller, with automatic takeover by another computer in case of failure.
  7. Data serving over OPC and EtherCAT (with ultra-low latency; 3.5 ms measured in current setup).
  8. Responsive live data display, real-time analysis, and storage on all nodes.
  9. Sustained, time-unlimited acquisitions (limited only by disk space).
  10. Offline analysis with data merging from different systems and tests.
  11. Specialized Display Node software with customizable views; unlimited display stations (33 in the IPRC system).
  12. FFT analysis in both online and offline modes.

Bustec designed and delivered to IPRC’s ICET Facility a fully synchronized, redundant data-acquisition system with roughly 15,000 channels, recording simultaneous, timestamped data — even across channels with different sampling rates. Multiple redundancy layers include dual power supplies and dual Ethernet ports on every ProDAQ instrument, plus mirrored data paths to EtherCAT and OPC UA servers. The result is a high-availability, high-integrity platform engineered to prevent data loss during mission-critical cryogenic engine and stage tests.

Bustec’s redundant DAQ system helped ISRO achieve reliable, synchronized data capture in one of the world’s most demanding test environments. Our mission is the same for every customer: deliver accuracy when it matters most.

At Bustec, we continue to develop high-accuracy, modular DAQ solutions for the most demanding environments. If you’d like to learn more about how our technology can support your testing requirements, get in touch with our team.