Hermasa launches AI unit for fish canning

Hermasa launches AI unit for fish canning

Hermasa has established a specialist business for canning production analytics. BlueAtlantIQ combines line monitoring, machine vision and artificial intelligence with machinery expertise developed for tuna and sardine processing.


IN Brief:

  • Hermasa Global Group has introduced BlueAtlantIQ Technologies alongside its established canning machinery business.
  • TunaView, vision systems and production analytics are intended to measure line efficiency and support process adjustments.
  • BlueAtlantIQ draws on Tecmopic, an approximately €2.5m research programme involving Hermasa and Anfaco-Cytma.

Hermasa has unveiled a specialist data and artificial intelligence business for the fish canning industry, creating BlueAtlantIQ Technologies as part of a new Hermasa Global Group structure. The company, based in Vigo, Spain, is bringing together machine monitoring, computer vision and production analytics developed through years of work on tuna-processing equipment.

The group comprises Hermasa Canning Technology, which designs, builds, installs and services equipment for tuna and sardine canning, and BlueAtlantIQ, which develops software and production data applications. Group president Sara Rodríguez Gude presented the structure in Vigo on 13 November. The separation gives the engineering business and the data operation distinct responsibilities while retaining their common industrial base in the city.

BlueAtlantIQ originates in an internal data team that began investigating applications of data science to canning in 2018. It is intended to connect operating information from equipment with production decisions, rather than requiring a processor to assess machinery performance through separate records after the shift has ended. Hermasa retains its engineering and industrial base in the Vigo area.

TunaView, one of BlueAtlantIQ’s products, displays productivity and yield data from tuna lines in real time. The business is also developing AI-assisted machine vision for inline product inspection and software to track overall equipment effectiveness by line and shift. Throughput data can reveal lost production time; inspection measurements can instead identify variation in the product passing through the machines. Both require sufficiently reliable data and a practical route for production teams to act on it.

The technology has developed alongside Tecmopic, a research programme carried out by Hermasa and the canning technology organisation Anfaco-Cytma since 2023. Its completion was marked in Vigo on 13 November. Work covered photonic sensing, near-infrared measurement, intelligent packing machinery and machine vision for tuna lines. Regional reporting puts the investment at approximately €2.5m, including about €1.3m in public support, and describes tests in operating factories involving 27 researchers.

Tests have investigated measurements of fish weight, moisture and liquid absorption as product moves through the line. The technical objective is to provide sufficiently frequent information to adjust packing operations while production continues. Measuring these properties presents a different challenge from counting finished cans: the raw material is variable, and packing equipment must account for the interaction between fish portions, added oil or other liquid, target fill weight and the speed of downstream operations.

Computer vision can classify visible characteristics of the product, while other sensors supply information not available from an ordinary image. Near-infrared instruments, for example, can be used to examine the optical response of a material at selected wavelengths. Any practical measurement depends on calibration, product presentation and the measurement conditions; it cannot simply be transferred from one raw material or machine to another without validation.

The Tecmopic trials have also used machine-learning models to propose and apply adjustments to packing machinery, according to the project’s reported results. Such corrections rely on measurements that distinguish a genuine deviation from normal raw material variation. Hermasa and its partners have not supplied an independently verified percentage improvement that can be applied to every installation, and performance will depend on the machinery, product and operating conditions.

Tuna arriving at a cannery varies in size, moisture content and condition, and those differences continue to affect packing after preparation and cooking. Fish portions, added oil and other filling liquids must be managed against the product specification while the line maintains its intended speed. A monitoring system needs to separate ordinary variation from process faults, otherwise additional sensors can increase the volume of alarms without improving the decisions taken on the line.

Overall equipment effectiveness provides another view of factory performance by relating availability, speed and quality. Tracking those dimensions by line and shift can reveal whether lost output is principally associated with stops, reduced running speed or rejected product. The calculation is only as useful as the definitions and downtime records behind it, so a consistent classification of stops and losses is needed before comparisons between shifts or plants are meaningful.

Hermasa’s existing machinery operation supplies equipment and complete lines across international canning markets. The separate software business gives it an organisational route to offer monitoring and optimisation alongside engineering support. The group has not disclosed customer installation figures, subscription revenues or a guaranteed yield improvement, leaving commercial deployment and measured performance as questions for future operations.

The new group plans to retain engineering and management functions in Vigo while developing BlueAtlantIQ’s applications for the international canning industry. It has not identified a confirmed installed customer base or published comparative yield results for the new software operation. Commercial operating data will establish the extent to which the trialled sensors, vision systems and control models can deliver repeatable improvements across different fish products and production lines.


Stories for you


  • Hermasa launches AI unit for fish canning

    Hermasa launches AI unit for fish canning

    Hermasa has established a specialist business for canning production analytics. BlueAtlantIQ combines line monitoring, machine vision and artificial intelligence with machinery expertise developed for tuna and sardine processing.


  • Buitoni consolidates US refrigerated pasta production

    Buitoni consolidates US refrigerated pasta production

    Buitoni is transferring refrigerated pasta manufacturing from Virginia to Massachusetts. Its Lawrence facility will absorb refrigerated pasta and sauce production, while Danville is scheduled to close on 28 December with 251 affected workers.