BRENNEN

Industries / Technology / Lasers and photonics

Cutting and welding

Laser processing is the use of a focused beam to drill, cut, weld or heat a material, from car body panels to the thin silicon film behind a phone screen.

5

Milestones achieved, from 1966 to 2025.

1

To watch for: announced, not yet confirmed outside.

2

Open deliverables. Nobody has shipped them.

The road

AchievedWatch forOpen

Achieved

  1. 1966

    Lasers pierce diamond dies

    Wire is made by pulling metal through a hole in a die, and the hardest-wearing dies are diamond. A 1966 article in Wire & Wire Products described how Western Electric used concentrated laser beams to pierce the holes in diamond wire-drawing dies, and said the method promised lower time and cost in die making.

    wirenet.org

    Credit

    Western Electric

  2. 1985

    Car panels welded before pressing

    A tailored blank is a sheet made by welding steel sheets of different thickness or grade together before the press shapes it, so that a car part is thick or strong only where it needs to be. A patent assigned to Littell International, a maker of blank-welding systems, says the first significant production line for car body panels was started by Thyssen Stahl in Germany in 1985 and used laser welding, after A. O. Smith had made tailored blanks with an electron beam in the US as early as 1967. Nippon Steel's 2003 technical report says the method lapsed after a 1967 side panel and was revived by Japanese and European carmakers around 1985 for side and floor panels, and that a laser weld bead is a little under 1 mm wide.

    patents.google.comnipponsteel.com

    Credit

    Thyssen Stahl

  3. 1986

    Fast silicon grown on glass

    Each pixel of a screen is switched by a thin-film transistor built on the glass, and transistors of amorphous (disordered) silicon switch slowly. T. Sameshima and S. Usui of Sony's research center in Yokohama melted thin silicon films with pulses of 308 nm ultraviolet light from a xenon chloride excimer laser, so that the film recrystallized as polysilicon without heating the glass, and made transistors with an electron mobility of 180 cm²/V·s at a process temperature of 260 °C. Excimer laser annealing now makes the low-temperature polysilicon (LTPS) behind OLED phone screens, with a line of laser light up to 1.5 m long, by Optics & Photonics News's 2026 account.

    LessonThe line beam has been hard to scale: Display Daily's Bob Raikes wrote in 2020 that LTPS plants had stopped at Generation 6 glass, about 1.5 m by 1.8 m, because the laser must be held precisely in focus, and that the difficulty of large LTPS was one reason Samsung's project to make large RGB OLED TVs failed.

    doi.orgoptica-opn.orgdisplaydaily.com

    Credit

    T. Sameshima and S. Usui, Sony

  4. 1997

    Pulses too short to heat

    Lasers with longer pulses heat the material around the spot they remove, which can damage the edge of a fine cut. X. Liu, D. Du and Gérard Mourou at the University of Michigan described how pulses lasting femtoseconds (millionths of a billionth of a second) remove material with a very small heated volume, and applied them to precision micromachining. Such ultrashort-pulse lasers are now established in display plants, where they cut phone cover glass into curved shapes with notches and camera holes, by Optics & Photonics News's 2026 account.

    api.openalex.orgoptica-opn.org

    Credit

    X. Liu, D. Du and Gérard Mourou, University of Michigan

  5. 2025

    Desktop engravers recalled

    Every laser product sold in the US must meet a federal performance standard enforced by the Food and Drug Administration (FDA). On 19 February 2025 Flux Technology of Taipei began recalling 91 Lazervida 10 W diode laser cutters and engravers sold in the US, because they lacked the beam attenuator, remote interlock connector, key switch, labels and manual instructions the rules require; the FDA posted the recall on 3 June 2025. In August 2024 Epilog Laser of Golden, Colorado, had begun recalling 751 Fusion Pro 24 engravers, rated as Class 2 products because the beam is enclosed, after some were built with an interlock in the wrong place, so the machine could run with its front door open and expose users to Class 4 laser light, the most hazardous class.

    Lessonan enclosed laser is only as safe as its interlock, and a misplaced switch turned a Class 2 machine into a Class 4 hazard.

    accessdata.fda.govaccessdata.fda.gov

    Credit

    the FDA's device recall records; Flux Technology; Epilog Laser

Watch for

  1. Glass cores drilled by laser

    Makers of the largest chip packages for AI plan to build them on glass cores, which need through-glass vias (TGVs), fine holes that carry connections from one side of the glass to the other. LPKF makes them with a laser-and-etching process it calls LIDE (laser induced deep etching). At NEPCON Japan on 22 January 2026 Intel showed a sample joining its EMIB packaging to a glass substrate and, as TrendForce reports, said it found no micro-cracks in testing; TrendForce expects Intel, TSMC, Samsung, Rapidus and SK's Absolics to begin mass production between 2027 and 2028.

    Achieved whena chip on a glass-core substrate ships in a product on sale, confirmed by the buyer or by a teardown.

    trendforce.comlide.lpkf.com

    Claimed by

    Intel, January 2026

Open

  1. 01

    Laser weld settings, gathered

    An open, cited table of published laser welding trials. Each row gives the material and thickness, the joint, the laser type and wavelength, the power, travel speed, spot size, focus position and shielding gas, and the result the paper reports: penetration depth, bead width and any pores or cracks. Every row points to the table or figure it came from. No laser is used; the work is reading papers.

    The brief
    Credit · unclaimed
  2. 02

    Laser fumes, listed by material

    An open, cited table of what laser cutting and engraving release from common shop and hobby materials, such as acrylic, plywood, MDF, cardboard, leather, polycarbonate and PVC: the gases found and the particle sizes and concentrations measured, each with its conditions (laser type and power, enclosure, extraction, time after the cut) and a link to the measurement. A second part records which school and makerspace safety lists ban or allow each material, and whether any measurement backs the rule. No laser is run; the work is reading studies and lists.

    The brief
    Credit · unclaimed

Open 01 · the brief

Laser weld settings, gathered

Credit · unclaimed

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

What to build

An open, cited table of published laser welding trials. Each row gives the material and thickness, the joint, the laser type and wavelength, the power, travel speed, spot size, focus position and shielding gas, and the result the paper reports: penetration depth, bead width and any pores or cracks. Every row points to the table or figure it came from. No laser is used; the work is reading papers.

Why it is still open

Researchers who predict weld quality from data write that supervised learning is "hindered in industrial applications by their dependence on extensive, high-quality labelled data" (Sen Li and colleagues, 2026). GitHub searches on 20 November 2026 for "laser welding dataset", "laser welding parameter" and "laser weld" found melt-pool image sets (ITMO's LaserWeldMonitor, MoltenPool2D) and simulation code (LaserbeamFoam), but no table of published settings and outcomes. The nearest compilation, a 2022 University of Pisa dataset, turns welding settings from the literature into dimensionless numbers but takes bead sizes from the groove sizes that ISO 9692 recommends, not from each paper's measured result.

arxiv.orgdata.mendeley.com

Built with what exists

OpenAlex and arXiv to find open-access papers; a Claude-based document reviewer that pulls values from tables and flags rows that do not match the page (see the Note "Seven more flagged"); Python with pandas; Zenodo to publish the table with a version number.

  1. First step

    Gather 20 open-access papers on fiber laser welding of 304 stainless steel and enter their trials by hand to settle the table's columns.

  2. Done looks like

    At least 1,000 trials from at least 150 papers, covering at least carbon steel, stainless steel, aluminum and copper, each value linked to its page; a second person checks a random 10% of rows and finds at least 98% correct; a simple model fitted to the table predicts penetration depth for trials from papers it was not fitted on, and its error is published.

Open 02 · the brief

Laser fumes, listed by material

Credit · unclaimed

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

What to build

An open, cited table of what laser cutting and engraving release from common shop and hobby materials, such as acrylic, plywood, MDF, cardboard, leather, polycarbonate and PVC: the gases found and the particle sizes and concentrations measured, each with its conditions (laser type and power, enclosure, extraction, time after the cut) and a link to the measurement. A second part records which school and makerspace safety lists ban or allow each material, and whether any measurement backs the rule. No laser is run; the work is reading studies and lists.

Why it is still open

Measurements are published one material at a time. A UCLA team measured CO2 laser cutting of acrylic in 2023: although many cutters have fume extractors, it found nanoplastic particles of 15.4 to 86 nm released each time the lid was opened, rising for at least 20 minutes after a cut. GitHub searches on 20 November 2026 for "laser cutter materials safety", "laser cutting materials list" and "laser cutter fumes" returned nothing.

api.openalex.org

Built with what exists

OpenAlex and Europe PMC to find studies; a small model built for extraction, such as Meta's Muse Ember, to pull values from papers for a person to check (see the Note "The jobs agents repeat"); Python with pandas; the safety lists that universities and makerspaces publish.

  1. First step

    Collect 10 university and makerspace lists of materials allowed in laser cutters and the first 10 emission studies, and lay out the table for acrylic, plywood and PVC.

  2. Done looks like

    At least 30 materials, each with at least one cited measurement and its conditions; at least 50 published safety lists compared; every material marked "measured", "rule without measurement" or "no data"; a second person checks a random 10% of values against the sources and finds none wrong.