TDK Bought OQmented's Laser Displays Without the Warranties

TDK announced on August 18 that it had bought assets from OQmented, a German developer of MEMS-based laser beam scanning displays. The release named no price and gave no closing date. TDK filled the gaps in afterward, telling The XR Beat that its AIsight unit had closed the transaction back in May, hired OQmented's team, and declined to carry over the German company's existing product support, warranties and customer obligations.

Customers of a display supplier found out in late August that their supplier's obligations had not survived the sale.

Three owners, two insolvencies

The corporate trail is short. OQmented GmbH entered formal insolvency proceedings on July 1, 2025, and kiiz Technologies GmbH took over its assets and its team that October, which is why TDK's press release names kiiz rather than OQmented as the entity it bought from. kiiz did not last either. Proceedings against it opened on April 1, 2026. TDK AIsight closed the month after.

Two insolvencies in nine months. One set of patents.

The money behind them was real. OQmented raised a $20 million Series A in 2023 with Sharp joining as a strategic investor, and it still was not enough, which is less an indictment of the company than a description of how this layer of the stack works. Upstream optics suppliers carry long development cycles and sell into a small number of device programs, so a single delayed headset or cancelled glasses project can erase most of a year of revenue. The same arithmetic surfaced when Magic Leap cut 193 jobs and repositioned itself as a components supplier rather than a headset maker.

What a light engine actually requires

A laser beam scanning display draws its picture by bouncing modulated light off a mirror a few millimeters across, moving it fast enough that the eye reads a continuous image rather than a travelling dot. OQmented built the mirror.

Laser scanning keeps reappearing in glasses designs for a structural reason. The engine only lights the pixels it draws, so an unlit pixel is genuinely off rather than dimly grey, and brightness scales with laser power instead of panel area. Area is the scarce resource in a temple arm. That combination is why the approach survives round after round of being declared impractical.

The mirror is not the product. A shippable light engine also needs lasers, drive electronics, calibration, packaging, thermal headroom and system integration, and every one of those is somewhere yield goes to die. TDK says it will pair the acquired work with its own full-color laser module, MEMS mirror technology, optics and eye-intent sensing, the company's term for reading where a wearer is looking and what they mean by it, to produce a complete display system under TDK AIsight. It also says it holds the rights needed to commercialize products built on what it bought.

It is the second purchase pointed this way. TDK acquired SoftEye in 2025 for custom chips, cameras, software and algorithms built for smart glasses, then stood up AIsight to hold the display business. A components maker is trying to sell a subsystem instead of a part. The pitch to a glasses manufacturer is that one supplier absorbs integration work that would otherwise land on its own engineers. XPANCEO and JBD are pushing the same way from the microLED side, and it is the reason xMEMS is selling active cooling into a category that never had room for a fan.

Glasses are not the only target. TDK's release lists automotive, industrial and mobility applications alongside consumer ones, and laser scanning already has a head-up display market to fall back on. Most waveguide startups cannot say that. It is a large part of why the technology was worth rescuing at all.

The part TDK has not shown

TDK declined to give The XR Beat qualification results, manufacturing yields, production volumes or named design wins. It would not say how many OQmented engineers transferred. It would not give a price. It would not say when anything built on the technology can be sampled or ordered. Existing customers are left waiting to hear which of their earlier integrations, if any, carry over.

Ferrite cores, cassette tapes, and now laser light engines is an unusual lineage, though the question at the end of it has not moved much: can the thing be built ten thousand times without drifting. TDK posted $16.6 billion in sales in fiscal 2026 and employs about 107,000 people, so it can fund a qualification cycle that neither OQmented nor kiiz lived long enough to finish.

For anyone designing glasses around laser scanning, the position after the sale is the position before it, minus a support contract. The patents are safe. The engineers are employed. Neither fact shows that a light engine can be ordered, qualified and built in volume, and until a sample or a named design win appears, what TDK has bought is a credible route to production rather than production itself.