BRENNEN

Industries / Industrial equipment and services / IoT and industrial automation

Sensors

Sensors and transmitters measure temperature, pressure, flow, level and vibration and send each reading on to a controller, usually as an electric signal on a pair of wires.

4

Milestones achieved, from 1975 to 2007.

1

To watch for: announced, not yet confirmed outside.

2

Open deliverables. Nobody has shipped them.

The road

AchievedWatch forOpen

Achieved

  1. 1975

    One signal for every instrument

    ISA's standard ANSI/ISA-50.00.01-1975 made a current between 4 and 20 milliamps (mA), its size standing for the reading, the common way for an instrument to send a measurement to a controller. ISA calls it one of the most widely used standards in industrial automation and says instruments using this analog signal "still far outnumber all fieldbus instruments combined."

    isa.org

    Credit

    ISA, committee ISA50

  2. 1986

    Digital data on one wire

    Rosemount introduced HART (Highway Addressable Remote Transducer) in 1986, the Oil & Gas Journal reported in 1991, when 18 companies formed a users' group to promote it. HART sends digital data at 1,200 bits per second on top of the 4–20 mA signal without interrupting it, so a plant can read and set up "smart" instruments over its existing wiring. FieldComm Group, which now looks after HART, counts more than 40 million HART instruments installed worldwide; it also dates HART's first release to the 1990s. Honeywell says it introduced its ST 3000 smart pressure transmitter, set up digitally over its signal wires with a handheld communicator, in 1983.

    ogj.comfieldcommgroup.orgpinnacle.yodify.com

    Credit

    Rosemount (HART); Honeywell (ST 3000)

  3. 2005

    Every level reading wrong

    On 23 March 2005 a distillation tower at BP's Texas City refinery was overfilled during startup, and the release that followed exploded, killing 15 people and injuring 180. The US Chemical Safety Board found that none of the tower's level instruments worked properly. The level transmitter was calibrated for a liquid of specific gravity 0.8, from a data sheet not updated since 1975, so with lighter hot liquid it showed 78% while liquid stood 158 ft (48 m) high; it could not read above 100% of its range. The sight glass was dirty and unreadable, and the backup high-level alarm never sounded.

    LessonAn instrument must be calibrated for the real fluid, tested from end to end, and backed by a second measurement that reads beyond the first one's range.

    csb.gov

    Credit

    US Chemical Safety Board (investigation)

  4. 2007

    Instruments without wires

    The HART Communication Foundation released WirelessHART in September 2007, a mesh radio network in which instruments relay one another's readings to a gateway and which works with existing HART devices. The IEC approved it as a publicly available specification, IEC/PAS 62591, on 19 September 2008, after a vote of national committees from 29 countries.

    controlglobal.com

    Credit

    HART Communication Foundation

Watch for

  1. Ethernet to the instrument

    Ethernet-APL instruments connect to the control system over a two-wire Ethernet cable in place of a 4–20 mA loop. ABB said that it and the mining company Boliden had deployed Ethernet-APL instruments at Boliden's zinc plant in Kokkola, Finland, "among the first installations of its kind in the country"; Boliden's engineer called it a chance to evaluate the technology in production. ARC Advisory Group wrote in 2026 that adoption is "still in the pilot-to-first-rollout phase."

    Achieved whenoutside reports show several plants running Ethernet-APL instruments in routine production, with instrument counts and fault rates.

    automation.comarcweb.com

    Claimed by

    ABB and Boliden, October 2026

Open

  1. 01

    Test wireless sensors from outside

    A published bench test of commercial wireless vibration sensors, the battery-powered units fixed to motors and pumps to warn of wear. Each sensor is mounted on a shaker table beside a calibrated reference accelerometer and shaken along its main axis and across it from 10 Hz to 1,000 Hz, while its current draw is logged to check the battery life its maker claims.

    The brief
    Credit · unclaimed
  2. 02

    Check settings against data sheets

    An AI agent that reads every transmitter's stored setup (its range, units and, for level instruments, the liquid density it assumes), compares it with the instrument's data sheet and the plant's process conditions, and lists mismatches like the one at Texas City for an instrument engineer to confirm.

    The brief
    Credit · unclaimed

Open 01 · the brief

Test wireless sensors from outside

Credit · unclaimed

The first to publish is credited on this line, and the deliverable moves from Open to Achieved.

What to build

A published bench test of commercial wireless vibration sensors, the battery-powered units fixed to motors and pumps to warn of wear. Each sensor is mounted on a shaker table beside a calibrated reference accelerometer and shaken along its main axis and across it from 10 Hz to 1,000 Hz, while its current draw is logged to check the battery life its maker claims.

Why it is still open

The tests published so far are makers' own. In January 2025 KCF Technologies, a sensor maker, tested a representative design on a shaker and estimated cross-axis errors of 60% to 410% for common designs from their data sheets, without naming them; it says many wireless sensor companies "simply ignore the accuracy of the overall sensor" and quote the accelerometer chip inside. A reliability trade article notes that battery life "can be loosely specified and difficult to test." Searches found no independent comparison of named products.

kcftech.comreliableplant.com

Built with what exists

an electrodynamic shaker and a reference accelerometer with a current calibration certificate, as found in university vibration labs; a bench meter that logs microamp currents; sensors bought from makers at list price; Python to fit the results.

  1. First step

    Buy two sensors from different makers and book time on a university lab's shaker to test them at 1 g from 10 Hz to 1,000 Hz.

  2. Done looks like

    Results for at least eight sensors from at least five makers, with raw data: amplitude error at each test frequency on the main and cross axes, and measured energy per reading turned into battery life at each maker's stated reporting interval.

Open 02 · the brief

Check settings against data sheets

Credit · unclaimed

The first to publish is credited on this line, and the deliverable moves from Open to Achieved.

What to build

An AI agent that reads every transmitter's stored setup (its range, units and, for level instruments, the liquid density it assumes), compares it with the instrument's data sheet and the plant's process conditions, and lists mismatches like the one at Texas City for an instrument engineer to confirm.

Why it is still open

The check is done by hand: calibration guidance tells technicians to verify the tag number and the lower and upper range values against the instrument data sheet with a handheld HART communicator. At Texas City the transmitter's data sheet had not been updated since 1975. Searches found no tool that audits instrument settings against process data automatically.

automationforum.cocsb.gov

Built with what exists

a USB HART modem, or the configuration export of asset-management software such as Emerson's AMS Device Manager; AI models that read PDF data sheets and tables (see the Note "Seven more flagged" for a model finding problems in long documents that people had missed); a spreadsheet of process conditions.

  1. First step

    Write 50 data sheets and matching configurations, plant ten errors, including a level transmitter set for the wrong density, and count how many a model finds.

  2. Done looks like

    On a test set of at least 300 instrument records with errors planted by an instrument engineer who keeps the answer key, the agent finds at least 90% of them with fewer than one false flag per 20 instruments, followed by one run on a real plant's records compared with a manual audit.