BLT DEEP DIVETechnology the market does not priceVol.1 · September 2026
SECTOR · BATTERY22
Samsung SDI’s Ulsan Line is Already in the Patent Record
What the KRW 4.45 trillion buys, and what “ultra-precision roll press control” is
By Karl Taekyun Chung, Partner Attorney and Head of Strategy, BLT Patent & Law Firm | karl@BLT.kr | +82-2-514-0104
Published September 4, 2026 · Data as of September 3, 2026
Key takeaways
1
Among the public records, the only structure corresponding to the “ultra-precision roll press control” the reports named is a compliant layer over a fluid chamber: a soft layer on the roll’s outer surface with fluid beneath it, pressing across a surface rather than along a line, with chamber depth limited as a function of electrode-plate thickness.
2
The equipment going into Ulsan is therefore unlikely to be a rigid roll with refined pressure control alone. For gauging what the KRW 4.45 trillion of capacity is specified to do, this specification is the most concrete clue available.
3
Process and equipment filings rose from 51% to 74% of the portfolio in a single year and named inventors from 211 to 266. Inspection and yield, however, slipped from 4.3% to 3.3%; whether that axis fills in is the practical indicator for the 2027 schedule.
On August 21, Samsung SDI disclosed the sale of its 33% stake in Samsung Display for KRW 4.45 trillion, earmarking the proceeds for U.S. production and all-solid-state batteries. Ten days earlier, press reports said it would replace the batch pressing used on the Suwon pilot line (S-Line) with roll pressing on the Ulsan mass-production line, naming “ultra-precision roll press control technology” as the decisive variable.
What that control actually is has never been stated. The company has not explained it; the reports carried only the name. This report fills that gap from the patent record. Two specifications filed in 2024 describe a structure that produces near-isostatic load with a roll, and the one process step it eliminates. The same record points to what that line does not yet have.
How we read the next 12–24 months
| | Outlook | Basis | Checkpoint |
| 1 | The Ulsan roll will carry a compliant layer over a fluid chamber, not a rigid surface | The specification gives a load “approaching an isostatic load” and limits chamber depth to a function of plate thickness | Suppliers’ order disclosures; Ulsan purchase orders |
| 2 | Inspection and yield filings keep increasing; if not, the schedule must be reconsidered | Filings naming inspection or determination: 4 in 2023, 7 in 2024. Both assume full inspection | Filings from H2 2025–2026, published in H1 2027 |
| 3 | Material and equipment sourcing consolidates around the current co-applicants | Three of four joint filings requested examination on filing; two are registered. The sole filings requested none | Supply and equity investment disclosures |
| 4 | The joint assets with Hyundai Motor Group stay on the isostatic side | Hyundai is first-named on the preheating WIP case; both roll-press cases are Samsung SDI alone | Filings published in H1 2027 |
In this issue
1Where all-solid-state stands todayThe chemistry is settled, the process is not
2Why “how you press it” is the gateSix devices, one problem
3What “ultra-precision roll press control” isA name in the press, a structure on file
4Making them is not enough — they must be sortedYield decides the schedule
5The partner list has changedA materials maker and an equipment maker join
6The center of gravity has moved to the factoryInspection and yield is still thin
1Where all-solid-state stands today
Korea’s three makers chose the same chemistry. What separates them is how they build it
All-solid-state batteries were long called the “dream battery”: replace the liquid electrolyte with a solid and the cell does not catch fire, and more energy fits in the same volume. The question was whether they could be built at all. By mid-decade that question changed. The industry now asks who gets there first, and at what cost.
Solid electrolytes divide into sulfide, oxide and polymer families. At a 2021 industry seminar Samsung SDI stated that “polymers are easy to process but lack durability, and oxides are difficult to mass-produce, so we chose sulfides.” LG Energy Solution, SK On and Toyota also chose sulfides. The route is therefore already settled across the major players. The remaining variable is not the material but the process.
Table 1. All-solid-state programs of the three Korean makers (public sources)
| | Target | Electrolyte | Process route | Pilot line |
| Samsung SDI | 2027 | Sulfide; lithium-precipitation (anode-free) type with a silver–carbon composite layer | Suwon S-Line WIP → Ulsan roll press | Suwon S-Line, running since March 2023 |
| SK On | 2029 | Sulfide | WIP-free (uniform pressing, 25–100℃) | Daejeon research center, completed September 2025 |
| LG Energy Solution | 2030 | Sulfide | Dry electrode | Ochang |
Source: company announcements and industry press through August 2026. Note: this table summarizes what each company has stated publicly. The patent analysis in this report is limited to Samsung SDI.
Samsung SDI leads on timing: its 2027 target is two years ahead of its domestic peers and the Suwon pilot line has run since March 2023. In December 2023 it created an ASB Commercialization Team inside its Battery Business Division dedicated to industrialization, and in October 2025 signed a three-way agreement under which it builds cells with Solid Power’s electrolyte and BMW validates them in vehicles.
The product took shape at InterBattery in March — an anode-free sulfide cell, around 900 Wh/L, in pouch format. The first commercial target is a humanoid robot, not an electric vehicle. Finished samples go to customers in the second half, with mass production planned for H2 2027.
That choice shows where the demand comes from. A humanoid carries roughly 2–4 kWh: Hyundai’s Atlas about 3.7 kWh, Tesla’s Optimus Gen2 and Figure 02 around 2.3 kWh. If more energy does not fit in the same volume, operating time does not increase. Energy storage pushes from the other side: after a series of fires in Korea, safety moved to the front of the purchasing criteria. All-solid-state sells on volumetric energy density and safety, and both come down to how the cell is pressed together.
The funding is in place. In July 2026 the company disclosed investment plans of KRW 9 trillion in Cheonan and KRW 16 trillion in Ulsan, and on August 21 it raised KRW 4.45 trillion by selling its Samsung Display stake. Ten days earlier, reports said batch WIP on the Suwon S-Line could not sustain mass-production throughput and that Ulsan would move to roll pressing, with “ultra-precision roll press control technology” named as the decisive variable. All of this the market already knows.
So what do the filings add? The reports carried the phrase “ultra-precision roll press control” without saying what that control is. Nor has the company explained it. What the roll looks like, who builds it, and what that line still lacks — those three things are written in the specifications.
2Why “how you press it” is the gate
Six filings cover six different devices, but one problem
The solid electrolyte layer sits between cathode and anode, and ions cross it only where it makes intimate contact with both plates. Incomplete contact becomes resistance, and current through that spot generates local heat. Life shortens, and the safety requirement described above is decided here. So all-solid-state manufacturing necessarily includes a pressing step — a stage that does not exist for liquid-electrolyte cells.
There are two ways to press. One is warm isostatic pressing (WIP), in which the cell is placed in a sealed vessel and a fluid applies equal pressure from every direction. Pressure uniformity is excellent, but it is a batch process: slow and hard to automate. The other is roll pressing, in which the stack passes between a pair of rolls. It runs continuously along a line and is fast, but a roll presses along a line of contact, so pressure is not uniform. Gain speed and you lose uniformity; keep uniformity and you lose speed.
Table 2. What the six claims actually limit
| Filing date | Applicants | What the claim limits |
| Nov 16, 2023 | MIRAE Company · Samsung SDI | A pressing pouch with perforated tear lines and fold lines so the packaging can be removed after pressing → handling in a batch process |
| Nov 22, 2023 | Samsung SDI · Hyundai Motor · Kia | A thermocouple measuring the pressing medium in an isostatic press, its tip protruding 5–20 mm from the inner cover surface → measurement precision in a batch process |
| Feb 26, 2024 | Samsung SDI (sole) | A reinforcing member placed beside the workpiece on the carrier film before passing through the pressing rolls → compensating the pressure drop at roll-press edges |
| Mar 28, 2024 | Hyundai Motor · Kia · Samsung SDI | Preheating the cell before WIP, with the preheat temperature set equal to the WIP process temperature → shorter cycle time in a batch process |
| Aug 20, 2024 | Samsung SDI (sole) | A compliant layer on the outer surface of the pressing roll over a fluid chamber, pressing by surface contact → near-isostatic uniformity in a continuous process |
| Dec 17, 2024 | MIRAE Company · Samsung SDI | A spherical joint on the pressing head with a tiltable body → self-alignment and uniform pressing even when the workpiece is inclined |
Source: claims of the published and registered Korean gazettes. Applicant order follows the gazette. Legal status is stated as of each gazette’s publication; subsequent examination requests or registrations were not checked.
The six describe entirely different devices: a pouch, a thermocouple, a reinforcing member, a preheating bath, a fluid chamber, a tilting head. As a list they look unrelated. But look at what each claim is actually trying to control and they all point to the same place: how evenly pressure reaches the whole face of the solid electrolyte layer. Device filings of this kind are appearing in a row, ahead of material filings. What this company is wrestling with now is not a substance but a factory.
Place the six in date order and a direction appears. The 2023 filings, and the March 2024 filing on which Hyundai and Kia are first-named applicants, aim to make the batch process more precise and easier to handle. The February and August 2024 filings, made by Samsung SDI alone, aim instead to obtain batch-level uniformity in a continuous process. Those two are the subject of the next section.
3What “ultra-precision roll press control” is
The reports carried the name; the specification describes a structure
The move to roll pressing is no longer news. An industry outlet reported it in June 2025, and the August 10, 2026 report added the reason — batch WIP cannot sustain throughput — and said Ulsan would use roll pressing. What it named as decisive was “ultra-precision roll press control technology that maintains uniform interfacial resistance while minimizing particle damage to the sulfide solid electrolyte.” No one wrote down what that control consists of. The specification the company filed with the Korean Intellectual Property Office in August 2024 does.
Figure 1. Pressing-process filings by filing year
Isostatic · WIP · hydrostatic
Roll press · rolling
Source: KIPRIS and WIPS ON, classified by BLT. Most filings from 2025 onward have not yet published, so those counts are understated; read the 2023-to-2024 shift rather than the absolute numbers.
KR 10-2024-0111579 “Roll press apparatus and method for an all-solid-state secondary battery” The specification opens by setting the two processes side by side. A roll presses along a line, so excessive load damages the cell structure and elongation differs with direction, delaminating layers; warm isostatic pressing gives even pressure but limited throughput. The invention states it will overcome both at once [0006]. Conventional roll pressing inserted a gasket between the two substrates; the added component created structural discontinuity, and the substrate wrinkled or the active material peeled at the cell’s edges and corners [0143].
The solution is a compliant layer on the roll’s outer surface with a fluid chamber beneath it. As pressing begins the layer deforms to follow the plate’s edge geometry and absorbs stress concentration, and because it is supported by fluid the plate receives a load “approaching an isostatic load” [0148]. If chamber depth exceeds half the combined plate thickness, the layer never reaches the roll core even at maximum compression [0156]. A layer thicker at the center deforms progressively outward [0158].
CLAIM 1 (working translation of the published Korean text)
a guide line conveying, in a superposed state, a first electrode plate in which an anode active material and a first solid electrolyte layer are stacked on an anode substrate and a second electrode plate in which a cathode active material and a second solid electrolyte layer are stacked on a cathode substrate;
guide rolls conveying along the guide line; and
a pair of pressing rolls that roll-press the first electrode plate and the second electrode plate conveyed in the superposed state through the guide line to form a unit cell,
wherein the pressing roll comprises a compliant layer on an outer surface thereof so as to press the first electrode plate and the second electrode plate by surface contact.
Published as KR 10-2026-0026979 · filed Aug 20, 2024 · 12 claims (2 independent) · no examination request as of publication · claims not final Among the public records, this is the only structure that corresponds to that name. It produces the effect of isostatic pressing with a roll. The phrasing is the company’s own. A rigid roll presses along a single tangent line, so pressure spikes there — precisely the particle damage the reports referred to. Fill a chamber with fluid beneath a soft layer and the roll conforms to the plate, transmitting pressure across a surface. Even pressure means even interfacial resistance.
The most striking part is elsewhere. The specification lists as an effect of this structure that it “eliminates the process of applying and removing the pouch that conventionally packaged the unit cell.” Confining the fluid inside the roll delivers an isostatic load without wrapping the cell. In the first row of Table 2, the November 2023 co-filing with MIRAE Company put tear lines into a pressing pouch so it could be removed more easily. Within a year the problem moved from removing the pouch well to not using one at all — one process step gone, a question of cost and takt time.
The specification also carries clues to the equipment specification. Cold water in the chamber gives cold rolling and hot water gives hot rolling, so one machine delivers both conditions. Together with chamber depth pinned to a function of plate thickness, this reads less like a concept than like a structure drawn against actual plate dimensions.
The other filing, made in February of the same year, addresses a different side effect of continuous processing. Feeding discrete sheets between rolls creates an impact at the edge, where thickness changes abruptly, and damages the plate.
KR 10-2024-0027095 “Roll press method and apparatus for an all-solid-state secondary battery” Feeding a workpiece on a carrier film between rolls creates an abrupt thickness difference between occupied and empty areas, and the impact at that boundary damages the battery or electrode [0006]. The invention places a reinforcing member beside the position the workpiece will occupy, filling the step, then passes both through the rolls together. It is formed of metal, ceramic, polymer or composite, and may be a dense foil or film or a porous material [0010–0011].
CLAIM 1 (working translation of the published Korean text)
a first step of arranging and supplying a reinforcing member on at least one side of an expected position on a carrier film at which a workpiece to form an all-solid-state secondary battery is to be arranged;
a second step of arranging the workpiece on one side of the reinforcing member; and
a third step of running the carrier film together with the workpiece between pressing rolls so that the pressing rolls press the workpiece and the reinforcing member.
Published as KR 10-2025-0130955 · filed Feb 26, 2024 · 20 claims (2 independent) · no examination request as of publication · claims not final By February 2024 roll pressing was already the premise and the work was on its side effects: adoption had moved past evaluation into design. And the August specification states directly that “a roll press mass-production process applying load with solid rolling rolls is under development.” That is the company’s own sentence, two years before the August 2026 reports.
Why uniform pressure leads to safety and to smaller cells
The two defects the specification names do more than scrap a cell. Where the active material peels or the substrate wrinkles at the plate edge, the interface lifts at that spot. A lifted spot has higher resistance, and current passing through it generates local heat. Pressing evenly means not creating such spots, and the safety energy storage demands is decided in this process, not in the material.
Compactness follows from another passage. The structure the company called “anode-free” in March is written there as a lithium-precipitation type anode. No anode active material is included at assembly; on charging, a lithium metal layer forms between the current collector and the anode coating and serves as the anode [0106–0107]. Removing an entire layer contributes directly to volumetric energy density. Thickness then changes with every cycle, which is why the specification adds a thin film on the current collector to “further planarize the deposition morphology of the lithium metal layer” [0115]. An uneven starting surface gives uneven deposition. The compliant layer and fluid chamber are therefore not only about speed: the safety energy storage demands and the compactness robotics demands meet at one point — how evenly the cell is pressed. This last connection is our reading of two passages side by side; the company has not linked them.
4Making them is not enough — you must sort them
Without yield, the mass-production date means nothing
The last gate is yield: however good the equipment, if a large share of cells must be discarded the cost does not work. As mass production approaches, inspection and determination filings begin to appear. Among Samsung SDI’s all-solid-state filings, those whose titles contain inspection, determination, measurement or verification number 4 in 2023 and 7 in 2024, first appearing in Q4 2023 (8 and 11 including abstracts).
KR 10-2023-0175954 “Method for determining acceptability of a cathode plate for an all-solid-state secondary battery” The invention screens acceptability at the plate stage, before a cell is built [0006]. Brightness of sample plates, EIS data and the performance of cells made from them are matched into a database; plates under evaluation are measured for brightness alone with a colorimeter and compared against the range seen when performance was normal [0007–0009]. Brightness, the specification states, reveals dispersion in the coated film [0012].
CLAIM 1 (working translation of the published Korean text)
a first step of matching brightness data of a sample electrode plate used as a cathode plate for an all-solid-state secondary battery, electrochemical impedance spectroscopy (EIS) data, and cell performance data, and storing them in a database;
a second step of measuring brightness data of an electrode plate under evaluation; and
a third step of comparing the brightness of the electrode plate under evaluation with the brightness range obtained when the cell performance data of the sample electrode plate was within a normal range, and determining acceptability of the electrode plate under evaluation.
Published as KR 10-2025-0086370 · filed Dec 6, 2023 · 15 claims (1 independent) · no examination request as of publication · claims not final EIS is accurate but takes minutes per sheet; reading brightness with a colorimeter is non-contact and takes seconds. The invention builds the correlation between the slow accurate measurement and the fast indirect one in advance, and on the line looks only at color.
KR 10-2024-0165956 “Inspection apparatus for an all-solid-state battery and inspection method using the same” Sulfide solid electrolytes react with moisture, and the invention turns that into a means of inspection [0006]. The battery is sealed in a chamber, an inspection substance is introduced, and the product of its reaction with the electrolyte is detected, determining whether the electrolyte is exposed [0009–0014].
CLAIM 1 (working translation of the published Korean text)
a receiving chamber having a receiving space therein, the receiving space being configured to receive an all-solid-state battery;
an inspection substance supply module connected to the receiving chamber and configured to supply an inspection substance into the receiving space; and
a detection module connected to the receiving chamber,
wherein the detection module is configured to detect a reaction substance generated by a reaction between a solid electrolyte in the all-solid-state battery and the inspection substance.
Published as KR 10-2026-0076452 · filed Nov 20, 2024 · 20 claims (2 independent) · no examination request as of publication · claims not final A known weakness of sulfides becomes a means of inspection. Reaction gas means that cell is leaking, so sealing is verified without disassembly, and the “customer reliability” listed among the effects points at customer qualification. One filing screens plates before a cell is built, the other sealing after: one full-inspection point before the process and one after — a design with no reason to exist in a pilot stage that gets by on sampling. Only two exist so far; how thin that is shows in the next section.
5The partner list has changed
Universities have been joined by a materials maker and an equipment maker
On next-generation development the 2026 half-year report says only that the company is “strengthening cooperation with overseas laboratories and with research institutes at home and abroad alongside its own R&D.” It does not say with whom, or on what. The list of co-applicants does. Three industrial partners now appear in a structure that previously held only universities and research institutes. A joint filing records with the patent office that two parties developed something together, evidencing both the existence and the content of the cooperation.
Table 3. Industrial partners appearing in all-solid-state joint filings
| Partner | What the company is (public sources) | What the joint filing covers | Prosecution speed |
| Kumho Petrochemical (011780 KS) | Korea’s largest synthetic rubber producer, moving into EV and battery materials | Binder for solid electrolyte film casting. The claims recite copolymers of non-polar aromatic vinyl and conjugated diene monomers — its core rubber chemistry | Examination requested on filing (Oct 2023) → registered (Jun 2026) |
| Solid Ionics (private) | All-solid-state materials specialist in Gangneung; Samyang invested in 2020 and 2024 and is second-largest shareholder. In February 2025 it announced sub-0.7 µm electrolyte particles, making lithium sulfide (Li₂S) by a dry process without hydrogen sulfide or organic solvents | Method for producing sulfide solid electrolyte: milling and spraying a precursor solution, then forming particles by thermal decomposition — suited to fine particles | Examination requested Dec 2024 Published, pending |
| MIRAE Company (049950 KQ) | KOSDAQ-listed leader in display processing equipment (edge grinders), since extended into semiconductor equipment and surgical robots. Declared entry into all-solid-state equipment in 2020 | Both filings sit around the pressing step: a pressing pouch with tear lines for removal, and a self-aligning tilting pressing head | Both requested examination on filing One registered in 14 months (Jan 2025) |
Source: published and registered Korean gazettes, company announcements and industry press. Other co-applicants include Sogang University (10 filings), Hanyang University (5), Yonsei University, POSTECH, the Korea Electronics Technology Institute and the Korea Electrotechnology Research Institute. All five joint filings with Hyundai Motor and Kia concern all-solid-state batteries.
What matters is that the three occupy different positions. Kumho Petrochemical is materials. Sulfide solid electrolytes react with polar solvents, which severely restricts the binders available. The synthetic rubber chemistry this company has worked on for decades is precisely the family that meets that condition, and the monomer composition in the claims shows it. Solid Ionics is feedstock. It makes lithium sulfide, the starting material for the electrolyte, and refines it to sub-0.7 µm particles. Smaller particles mean more contact area with the cathode active material, which is the same objective as the pressing problem, approached from the materials side. MIRAE Company is equipment — and both of its filings sit around the pressing step. Both are recorded in the gazettes as outputs of a national R&D program of the Ministry of Trade, Industry and Energy, “Development of manufacturing equipment for all-solid-state battery cells of 50 cm² and above” (project no. 20012349), in 2023 and again in 2024.
Prosecution speed is also notable. Large companies usually defer examination requests while they watch the market. Yet among the joint filings with these three, three of four requested examination on the filing date, and two are already registered. By contrast, the four Samsung SDI sole filings discussed above and the two joint filings with Hyundai Motor Group carry no examination request as of publication. The timing of an examination request is a fact recorded in the gazette and its reasons are not disclosed, but it reads as an intent to convert the materials and equipment relationships into settled rights quickly.
6The center of gravity has moved to the factory
One axis, inspection and yield, is still thin
As a technology matures the kind of patent changes. While the question is what to make it from, material and composition filings appear; while the form is being fixed, electrode and cell structure filings appear. Only when how to make it is being settled do process and equipment filings pour out. Portfolio composition therefore shows the development stage independently of what a company announces.
Table 4. Composition of all-solid-state filings by character
| Character of the filing | Filed 2023 | Filed 2024 | Change |
| Electrolyte material and composition | 10.3% | 5.1% | ▼ 5.2pp |
| Electrode design | 27.7% | 15.7% | ▼ 12.0pp |
| Cell structure | 3.3% | 1.8% | ▼ 1.5pp |
| Manufacturing process | 46.7% | 58.6% | ▲ 11.9pp |
| Manufacturing equipment | 4.3% | 15.1% | ▲ 10.8pp |
| Inspection and evaluation | 4.3% | 3.3% | ▼ 1.0pp |
| Process and equipment combined | 51.0% | 73.7% | ▲ 22.7pp |
Source: KIPRIS and WIPS ON, classified by BLT. Each filing is assigned to a single axis, judged from its title and abstract, starting from the most downstream axis. The classification rules are set out in Appendix A. Base: 184 filings in 2023 and 331 in 2024. Six 2023 filings and one 2024 filing fall on no axis and are excluded, so the columns do not sum to 100%. Filings from 2025 onward are excluded because many have not published.
In one year the share taken by process and equipment rose by 23 percentage points. Over the same period the company created the ASB Commercialization Team inside its Battery Business Division (December 2023), and the number of inventors named in this field rose from 211 to 266, of whom 136 are new. Three separate records — organization, headcount and filings — point the same way at the same time. This is a company preparing a factory, not running a research program.
One place remains thin. While process and equipment climbed into double digits, inspection and evaluation fell from 4.3% to 3.3%. In absolute terms it rose from 8 filings to 11, so it has not gone backward; the other axes simply grew faster. But this axis usually thickens sharply in the one or two years before mass production. Given that both inspection filings already assume full inspection, it is more reasonable to read this axis as having just begun.
Outlook and checkpoints
| | Outlook | What confirms it | What would prove it wrong |
| 1 | The Ulsan roll will carry a compliant layer over a fluid chamber, not a rigid surface | Suppliers’ order disclosures; Ulsan purchase orders | If the order is placed for a plain rigid roll with only refined pressure control |
| 2 | Inspection and yield filings keep increasing; if not, the schedule must be reconsidered | Filings from H2 2025–2026, published in H1 2027 | If inspection know-how is kept as trade secret, or filed under the equipment maker’s name |
| 3 | Material and equipment sourcing consolidates around the current co-applicants | Supply agreements and equity investments; each party’s subsequent sole filings | If the joint filings end with a single funded project and nothing follows |
| 4 | The joint assets with Hyundai Motor Group stay on the isostatic side | Technical fields of joint filings published in H1 2027 | If a roll-press joint filing appears in Hyundai Motor Group’s name |
AData and method
Why the specification is read first
Korean applications filed by Samsung SDI Co., Ltd. and published through August 2026 whose
specifications contain the terms all-solid-state, solid electrolyte or solid secondary battery, filed between
January 2023 and July 2026. For the ten documents cited, the full gazettes were obtained and the
description and claims read directly
Why the specification is read first
Technical content is taken from the description; what the company seeks to monopolize is taken from the claims. Claims in applications that have not completed examination are drafted broadly and do not show the actual implementation, whereas the problem addressed and the effects stated record what the company treated as the problem. The narrative therefore rests on the description, with claim 1 reproduced so the two can be told apart.
Filings are classified from title and abstract into six axes — electrolyte material and composition, electrode design, cell structure, manufacturing process, manufacturing equipment, and inspection and evaluation — one axis per filing, taking the most downstream axis where several apply. A method of manufacturing an electrode counts as manufacturing process. This classification and every figure are recomputed by a single script and checked against the source data at review.
Limitations to read alongside
Korean applications publish 18 months from the earliest priority date, so most 2025-onward filings are
absent and year-on-year counts are meaningful only through 2024. Apart from the two registered cases,
none of the applications cited has final claims and six carried no examination request as of publication; the
claims reproduced are the applicant’s opening text and the final scope may differ. Technology kept as trade
secret is outside this analysis, and filing counts do not measure technical capability. Connecting the phrase
used in the press with the structure in the specification is our reading of public documents; the company
has not stated the two are the same. Korean filings only.
Terms
Warm isostatic pressing (WIP) — a batch method applying equal pressure from all directions with a fluid in a
sealed vessel. Roll pressing — a continuous method passing the stack between a pair of rolls. Line versus
surface pressure — the force of a roll pressing along a line of contact, versus force distributed over an area.
Lithium-precipitation (anode-free) anode — a structure with no anode active material at assembly, in which
lithium deposits on charging and serves as the anode. Interfacial resistance — resistance arising at the surface
where the solid electrolyte layer meets the electrode plate
Patent documents cited
| Application no. | Published / registered | Title | Applicants |
| 10-2023-0137478 | Reg. 10-2982143 | Binder for solid electrolyte film casting | Kumho Petrochemical · Samsung SDI |
| 10-2023-0159338 | Reg. 10-2753699 | All-solid-state battery and method of manufacturing the same | MIRAE Company ·
Samsung SDI |
| 10-2023-0163640 | Pub. 10-2025-0076213 | Isostatic pressing apparatus for an all-solid-state secondary battery | Samsung SDI ·
Hyundai Motor · Kia |
| 10-2023-0175954 | Pub. 10-2025-0086370 | Method for determining acceptability of a cathode plate | Samsung SDI |
| 10-2024-0004850 | Pub. 10-2025-0109987 | Method for producing sulfide solid electrolyte | Samsung SDI · Solid Ionics |
| 10-2024-0027095 | Pub. 10-2025-0130955 | Roll press method and apparatus | Samsung SDI |
| 10-2024-0042464 | Pub. 10-2025-0145265 | Preheating apparatus and manufacturing method including a preheating process | Hyundai Motor · Kia · Samsung SDI |
| 10-2024-0111579 | Pub. 10-2026-0026979 | Roll press apparatus and method | Samsung SDI |
| 10-2024-0165956 | Pub. 10-2026-0076452 | Inspection apparatus and inspection method using the same | Samsung SDI |
| 10-2024-0189236 | Pub. 10-2026-0096601 | Apparatus for manufacturing an all-solid-state battery | MIRAE Company ·
Samsung SDI |
Korean application, publication and registration numbers. Titles are working translations.
Other public sources
Samsung SDI half-year report (Aug 14, 2026); Samsung Display stake sale disclosure and related coverage (Aug 21, 2026); Digital Daily, “Samsung SDI’s all-solid-state moves beyond slogans” (Aug 10, 2026); Samsung SDI Newsroom and the InterBattery 2026 presentation (March 2026); disclosures of future business and management plans (Jul 2–3, 2026); SNE Research NGBS 2021 seminar (May 25, 2021); Electronic Times (Jun 5, 2025); partner announcements and press releases.
Disclaimer
This report is an industry analysis prepared by BLT Patent and Law Firm from public intellectual property databases, regulatory disclosures and press materials. It does not constitute investment advice or a recommendation to buy, sell or hold any security. Investment decisions remain the sole responsibility of the reader.
This report is not legal advice on any specific matter. The facts stated rest on materials publicly available as of the data date and may change with subsequent examination, registration or litigation. Claims may be amended during prosecution.
Companies mentioned may include clients of BLT Patent and Law Firm. This report was prepared using public data only; no confidential client information was used.
This report was not provided in advance to institutional investors or any third party. The author holds no shares in the companies mentioned and received no consideration from them.
Copyright belongs to BLT Patent and Law Firm. Quotation with attribution is permitted; reproduction or commercial use of the whole is not.
 | BLT DEEP DIVE Technology the Market Does Not See, and the Rights Behind It
‘BLT DEEP DIVE’ draws on publicly available patent and industry data to track changes in corporate technology and the business strategies behind them.
Published on the first Friday of every month, the full report is available free of charge on our website.
For individual advice on the technical areas covered in this issue, please contact us using the details below.
|
For more information, please contact BLT Patent and Law Firm. +82-2-514-0104 | info@BLT.kr | www.BLT.kr
| About the Author Taekyun Chung, Partner Patent Attorney, serves as Head of Strategy at BLT Patent and Law Firm, where he advises startups on IP strategy, business model (BM) strategy, and go-to-market (GTM) strategy.
He graduated from Yonsei University with a degree in Electrical and Electronic Engineering and passed the 48th Korean Patent Attorney Examination in 2011. He currently works with startups across a wide range of industries, supporting their growth not only through intellectual property matters—including patents,
trademarks, and designs—but also through active involvement in their business strategy and development. |
|
|
#SamsungSDI #AllSolidStateBattery #SolidStateBattery #AllSolidStateBatteries #BatteryTechnology #BatteryIndustry #NextGenerationBattery #AdvancedBattery #BatteryInnovation #EnergyStorage #BatteryManufacturing #BatteryProduction #BatteryManufacturingProcess #BatteryEquipment #RollPress #RollPressing #RollPressTechnology #WarmIsostaticPressing #WIP #SulfideSolidElectrolyte #SulfideBattery #AnodeFreeBattery #BatteryPatent #BatteryPatents #PatentAnalysis #PatentStrategy #PatentIntelligence #IPStrategy #IntellectualProperty #BLTDeepDive
Samsung SDI’s Ulsan Line is Already in the Patent Record
What the KRW 4.45 trillion buys, and what “ultra-precision roll press control” is
Published September 4, 2026 · Data as of September 3, 2026
On August 21, Samsung SDI disclosed the sale of its 33% stake in Samsung Display for KRW 4.45 trillion, earmarking the proceeds for U.S. production and all-solid-state batteries. Ten days earlier, press reports said it would replace the batch pressing used on the Suwon pilot line (S-Line) with roll pressing on the Ulsan mass-production line, naming “ultra-precision roll press control technology” as the decisive variable.
What that control actually is has never been stated. The company has not explained it; the reports carried only the name. This report fills that gap from the patent record. Two specifications filed in 2024 describe a structure that produces near-isostatic load with a roll, and the one process step it eliminates. The same record points to what that line does not yet have.
1Where all-solid-state stands today
Korea’s three makers chose the same chemistry. What separates them is how they build it
All-solid-state batteries were long called the “dream battery”: replace the liquid electrolyte with a solid and the cell does not catch fire, and more energy fits in the same volume. The question was whether they could be built at all. By mid-decade that question changed. The industry now asks who gets there first, and at what cost.
Solid electrolytes divide into sulfide, oxide and polymer families. At a 2021 industry seminar Samsung SDI stated that “polymers are easy to process but lack durability, and oxides are difficult to mass-produce, so we chose sulfides.” LG Energy Solution, SK On and Toyota also chose sulfides. The route is therefore already settled across the major players. The remaining variable is not the material but the process.
Source: company announcements and industry press through August 2026. Note: this table summarizes what each company has stated publicly. The patent analysis in this report is limited to Samsung SDI.
Samsung SDI leads on timing: its 2027 target is two years ahead of its domestic peers and the Suwon pilot line has run since March 2023. In December 2023 it created an ASB Commercialization Team inside its Battery Business Division dedicated to industrialization, and in October 2025 signed a three-way agreement under which it builds cells with Solid Power’s electrolyte and BMW validates them in vehicles.
The product took shape at InterBattery in March — an anode-free sulfide cell, around 900 Wh/L, in pouch format. The first commercial target is a humanoid robot, not an electric vehicle. Finished samples go to customers in the second half, with mass production planned for H2 2027.
That choice shows where the demand comes from. A humanoid carries roughly 2–4 kWh: Hyundai’s Atlas about 3.7 kWh, Tesla’s Optimus Gen2 and Figure 02 around 2.3 kWh. If more energy does not fit in the same volume, operating time does not increase. Energy storage pushes from the other side: after a series of fires in Korea, safety moved to the front of the purchasing criteria. All-solid-state sells on volumetric energy density and safety, and both come down to how the cell is pressed together.
The funding is in place. In July 2026 the company disclosed investment plans of KRW 9 trillion in Cheonan and KRW 16 trillion in Ulsan, and on August 21 it raised KRW 4.45 trillion by selling its Samsung Display stake. Ten days earlier, reports said batch WIP on the Suwon S-Line could not sustain mass-production throughput and that Ulsan would move to roll pressing, with “ultra-precision roll press control technology” named as the decisive variable. All of this the market already knows.
So what do the filings add? The reports carried the phrase “ultra-precision roll press control” without saying what that control is. Nor has the company explained it. What the roll looks like, who builds it, and what that line still lacks — those three things are written in the specifications.
2Why “how you press it” is the gate
Six filings cover six different devices, but one problem
The solid electrolyte layer sits between cathode and anode, and ions cross it only where it makes intimate contact with both plates. Incomplete contact becomes resistance, and current through that spot generates local heat. Life shortens, and the safety requirement described above is decided here. So all-solid-state manufacturing necessarily includes a pressing step — a stage that does not exist for liquid-electrolyte cells.
There are two ways to press. One is warm isostatic pressing (WIP), in which the cell is placed in a sealed vessel and a fluid applies equal pressure from every direction. Pressure uniformity is excellent, but it is a batch process: slow and hard to automate. The other is roll pressing, in which the stack passes between a pair of rolls. It runs continuously along a line and is fast, but a roll presses along a line of contact, so pressure is not uniform. Gain speed and you lose uniformity; keep uniformity and you lose speed.
Source: claims of the published and registered Korean gazettes. Applicant order follows the gazette. Legal status is stated as of each gazette’s publication; subsequent examination requests or registrations were not checked.
The six describe entirely different devices: a pouch, a thermocouple, a reinforcing member, a preheating bath, a fluid chamber, a tilting head. As a list they look unrelated. But look at what each claim is actually trying to control and they all point to the same place: how evenly pressure reaches the whole face of the solid electrolyte layer. Device filings of this kind are appearing in a row, ahead of material filings. What this company is wrestling with now is not a substance but a factory.
Place the six in date order and a direction appears. The 2023 filings, and the March 2024 filing on which Hyundai and Kia are first-named applicants, aim to make the batch process more precise and easier to handle. The February and August 2024 filings, made by Samsung SDI alone, aim instead to obtain batch-level uniformity in a continuous process. Those two are the subject of the next section.
3What “ultra-precision roll press control” is
The reports carried the name; the specification describes a structure
The move to roll pressing is no longer news. An industry outlet reported it in June 2025, and the August 10, 2026 report added the reason — batch WIP cannot sustain throughput — and said Ulsan would use roll pressing. What it named as decisive was “ultra-precision roll press control technology that maintains uniform interfacial resistance while minimizing particle damage to the sulfide solid electrolyte.” No one wrote down what that control consists of. The specification the company filed with the Korean Intellectual Property Office in August 2024 does.
Source: KIPRIS and WIPS ON, classified by BLT. Most filings from 2025 onward have not yet published, so those counts are understated; read the 2023-to-2024 shift rather than the absolute numbers.
The specification opens by setting the two processes side by side. A roll presses along a line, so excessive load damages the cell structure and elongation differs with direction, delaminating layers; warm isostatic pressing gives even pressure but limited throughput. The invention states it will overcome both at once [0006]. Conventional roll pressing inserted a gasket between the two substrates; the added component created structural discontinuity, and the substrate wrinkled or the active material peeled at the cell’s edges and corners [0143].
The solution is a compliant layer on the roll’s outer surface with a fluid chamber beneath it. As pressing begins the layer deforms to follow the plate’s edge geometry and absorbs stress concentration, and because it is supported by fluid the plate receives a load “approaching an isostatic load” [0148]. If chamber depth exceeds half the combined plate thickness, the layer never reaches the roll core even at maximum compression [0156]. A layer thicker at the center deforms progressively outward [0158].
Among the public records, this is the only structure that corresponds to that name. It produces the effect of isostatic pressing with a roll. The phrasing is the company’s own. A rigid roll presses along a single tangent line, so pressure spikes there — precisely the particle damage the reports referred to. Fill a chamber with fluid beneath a soft layer and the roll conforms to the plate, transmitting pressure across a surface. Even pressure means even interfacial resistance.
The most striking part is elsewhere. The specification lists as an effect of this structure that it “eliminates the process of applying and removing the pouch that conventionally packaged the unit cell.” Confining the fluid inside the roll delivers an isostatic load without wrapping the cell. In the first row of Table 2, the November 2023 co-filing with MIRAE Company put tear lines into a pressing pouch so it could be removed more easily. Within a year the problem moved from removing the pouch well to not using one at all — one process step gone, a question of cost and takt time.
The specification also carries clues to the equipment specification. Cold water in the chamber gives cold rolling and hot water gives hot rolling, so one machine delivers both conditions. Together with chamber depth pinned to a function of plate thickness, this reads less like a concept than like a structure drawn against actual plate dimensions.
The other filing, made in February of the same year, addresses a different side effect of continuous processing. Feeding discrete sheets between rolls creates an impact at the edge, where thickness changes abruptly, and damages the plate.
Feeding a workpiece on a carrier film between rolls creates an abrupt thickness difference between occupied and empty areas, and the impact at that boundary damages the battery or electrode [0006]. The invention places a reinforcing member beside the position the workpiece will occupy, filling the step, then passes both through the rolls together. It is formed of metal, ceramic, polymer or composite, and may be a dense foil or film or a porous material [0010–0011].
By February 2024 roll pressing was already the premise and the work was on its side effects: adoption had moved past evaluation into design. And the August specification states directly that “a roll press mass-production process applying load with solid rolling rolls is under development.” That is the company’s own sentence, two years before the August 2026 reports.
Why uniform pressure leads to safety and to smaller cells
The two defects the specification names do more than scrap a cell. Where the active material peels or the substrate wrinkles at the plate edge, the interface lifts at that spot. A lifted spot has higher resistance, and current passing through it generates local heat. Pressing evenly means not creating such spots, and the safety energy storage demands is decided in this process, not in the material.
Compactness follows from another passage. The structure the company called “anode-free” in March is written there as a lithium-precipitation type anode. No anode active material is included at assembly; on charging, a lithium metal layer forms between the current collector and the anode coating and serves as the anode [0106–0107]. Removing an entire layer contributes directly to volumetric energy density. Thickness then changes with every cycle, which is why the specification adds a thin film on the current collector to “further planarize the deposition morphology of the lithium metal layer” [0115]. An uneven starting surface gives uneven deposition. The compliant layer and fluid chamber are therefore not only about speed: the safety energy storage demands and the compactness robotics demands meet at one point — how evenly the cell is pressed. This last connection is our reading of two passages side by side; the company has not linked them.
4Making them is not enough — you must sort them
Without yield, the mass-production date means nothing
The last gate is yield: however good the equipment, if a large share of cells must be discarded the cost does not work. As mass production approaches, inspection and determination filings begin to appear. Among Samsung SDI’s all-solid-state filings, those whose titles contain inspection, determination, measurement or verification number 4 in 2023 and 7 in 2024, first appearing in Q4 2023 (8 and 11 including abstracts).
The invention screens acceptability at the plate stage, before a cell is built [0006]. Brightness of sample plates, EIS data and the performance of cells made from them are matched into a database; plates under evaluation are measured for brightness alone with a colorimeter and compared against the range seen when performance was normal [0007–0009]. Brightness, the specification states, reveals dispersion in the coated film [0012].
EIS is accurate but takes minutes per sheet; reading brightness with a colorimeter is non-contact and takes seconds. The invention builds the correlation between the slow accurate measurement and the fast indirect one in advance, and on the line looks only at color.
Sulfide solid electrolytes react with moisture, and the invention turns that into a means of inspection [0006]. The battery is sealed in a chamber, an inspection substance is introduced, and the product of its reaction with the electrolyte is detected, determining whether the electrolyte is exposed [0009–0014].
A known weakness of sulfides becomes a means of inspection. Reaction gas means that cell is leaking, so sealing is verified without disassembly, and the “customer reliability” listed among the effects points at customer qualification. One filing screens plates before a cell is built, the other sealing after: one full-inspection point before the process and one after — a design with no reason to exist in a pilot stage that gets by on sampling. Only two exist so far; how thin that is shows in the next section.
5The partner list has changed
Universities have been joined by a materials maker and an equipment maker
On next-generation development the 2026 half-year report says only that the company is “strengthening cooperation with overseas laboratories and with research institutes at home and abroad alongside its own R&D.” It does not say with whom, or on what. The list of co-applicants does. Three industrial partners now appear in a structure that previously held only universities and research institutes. A joint filing records with the patent office that two parties developed something together, evidencing both the existence and the content of the cooperation.
(011780 KS)
(private)
Published, pending
(049950 KQ)
One registered in 14 months (Jan 2025)
Source: published and registered Korean gazettes, company announcements and industry press. Other co-applicants include Sogang University (10 filings), Hanyang University (5), Yonsei University, POSTECH, the Korea Electronics Technology Institute and the Korea Electrotechnology Research Institute. All five joint filings with Hyundai Motor and Kia concern all-solid-state batteries.
What matters is that the three occupy different positions. Kumho Petrochemical is materials. Sulfide solid electrolytes react with polar solvents, which severely restricts the binders available. The synthetic rubber chemistry this company has worked on for decades is precisely the family that meets that condition, and the monomer composition in the claims shows it. Solid Ionics is feedstock. It makes lithium sulfide, the starting material for the electrolyte, and refines it to sub-0.7 µm particles. Smaller particles mean more contact area with the cathode active material, which is the same objective as the pressing problem, approached from the materials side. MIRAE Company is equipment — and both of its filings sit around the pressing step. Both are recorded in the gazettes as outputs of a national R&D program of the Ministry of Trade, Industry and Energy, “Development of manufacturing equipment for all-solid-state battery cells of 50 cm² and above” (project no. 20012349), in 2023 and again in 2024.
Prosecution speed is also notable. Large companies usually defer examination requests while they watch the market. Yet among the joint filings with these three, three of four requested examination on the filing date, and two are already registered. By contrast, the four Samsung SDI sole filings discussed above and the two joint filings with Hyundai Motor Group carry no examination request as of publication. The timing of an examination request is a fact recorded in the gazette and its reasons are not disclosed, but it reads as an intent to convert the materials and equipment relationships into settled rights quickly.
6The center of gravity has moved to the factory
One axis, inspection and yield, is still thin
As a technology matures the kind of patent changes. While the question is what to make it from, material and composition filings appear; while the form is being fixed, electrode and cell structure filings appear. Only when how to make it is being settled do process and equipment filings pour out. Portfolio composition therefore shows the development stage independently of what a company announces.
Source: KIPRIS and WIPS ON, classified by BLT. Each filing is assigned to a single axis, judged from its title and abstract, starting from the most downstream axis. The classification rules are set out in Appendix A. Base: 184 filings in 2023 and 331 in 2024. Six 2023 filings and one 2024 filing fall on no axis and are excluded, so the columns do not sum to 100%. Filings from 2025 onward are excluded because many have not published.
In one year the share taken by process and equipment rose by 23 percentage points. Over the same period the company created the ASB Commercialization Team inside its Battery Business Division (December 2023), and the number of inventors named in this field rose from 211 to 266, of whom 136 are new. Three separate records — organization, headcount and filings — point the same way at the same time. This is a company preparing a factory, not running a research program.
One place remains thin. While process and equipment climbed into double digits, inspection and evaluation fell from 4.3% to 3.3%. In absolute terms it rose from 8 filings to 11, so it has not gone backward; the other axes simply grew faster. But this axis usually thickens sharply in the one or two years before mass production. Given that both inspection filings already assume full inspection, it is more reasonable to read this axis as having just begun.
Outlook and checkpoints
AData and method
Why the specification is read first
Korean applications filed by Samsung SDI Co., Ltd. and published through August 2026 whose
specifications contain the terms all-solid-state, solid electrolyte or solid secondary battery, filed between
January 2023 and July 2026. For the ten documents cited, the full gazettes were obtained and the
description and claims read directly
Why the specification is read first
Technical content is taken from the description; what the company seeks to monopolize is taken from the claims. Claims in applications that have not completed examination are drafted broadly and do not show the actual implementation, whereas the problem addressed and the effects stated record what the company treated as the problem. The narrative therefore rests on the description, with claim 1 reproduced so the two can be told apart.
Filings are classified from title and abstract into six axes — electrolyte material and composition, electrode design, cell structure, manufacturing process, manufacturing equipment, and inspection and evaluation — one axis per filing, taking the most downstream axis where several apply. A method of manufacturing an electrode counts as manufacturing process. This classification and every figure are recomputed by a single script and checked against the source data at review.
Limitations to read alongside
Korean applications publish 18 months from the earliest priority date, so most 2025-onward filings are
absent and year-on-year counts are meaningful only through 2024. Apart from the two registered cases,
none of the applications cited has final claims and six carried no examination request as of publication; the
claims reproduced are the applicant’s opening text and the final scope may differ. Technology kept as trade
secret is outside this analysis, and filing counts do not measure technical capability. Connecting the phrase
used in the press with the structure in the specification is our reading of public documents; the company
has not stated the two are the same. Korean filings only.
Terms
Warm isostatic pressing (WIP) — a batch method applying equal pressure from all directions with a fluid in a
sealed vessel. Roll pressing — a continuous method passing the stack between a pair of rolls. Line versus
surface pressure — the force of a roll pressing along a line of contact, versus force distributed over an area.
Lithium-precipitation (anode-free) anode — a structure with no anode active material at assembly, in which
lithium deposits on charging and serves as the anode. Interfacial resistance — resistance arising at the surface
where the solid electrolyte layer meets the electrode plate
Patent documents cited
Samsung SDI
Hyundai Motor · Kia
Samsung SDI
Korean application, publication and registration numbers. Titles are working translations.
Other public sources
Samsung SDI half-year report (Aug 14, 2026); Samsung Display stake sale disclosure and related coverage (Aug 21, 2026); Digital Daily, “Samsung SDI’s all-solid-state moves beyond slogans” (Aug 10, 2026); Samsung SDI Newsroom and the InterBattery 2026 presentation (March 2026); disclosures of future business and management plans (Jul 2–3, 2026); SNE Research NGBS 2021 seminar (May 25, 2021); Electronic Times (Jun 5, 2025); partner announcements and press releases.
Disclaimer
This report is an industry analysis prepared by BLT Patent and Law Firm from public intellectual property databases, regulatory disclosures and press materials. It does not constitute investment advice or a recommendation to buy, sell or hold any security. Investment decisions remain the sole responsibility of the reader.
This report is not legal advice on any specific matter. The facts stated rest on materials publicly available as of the data date and may change with subsequent examination, registration or litigation. Claims may be amended during prosecution.
Companies mentioned may include clients of BLT Patent and Law Firm. This report was prepared using public data only; no confidential client information was used.
This report was not provided in advance to institutional investors or any third party. The author holds no shares in the companies mentioned and received no consideration from them.
Copyright belongs to BLT Patent and Law Firm. Quotation with attribution is permitted; reproduction or commercial use of the whole is not.
BLT DEEP DIVE
Technology the Market Does Not See, and the Rights Behind It
‘BLT DEEP DIVE’ draws on publicly available patent and industry data to track changes in corporate technology and the business strategies behind them.
Published on the first Friday of every month, the full report is available free of charge on our website.
For individual advice on the technical areas covered in this issue, please contact us using the details below.
For more information, please contact BLT Patent and Law Firm.
+82-2-514-0104 | info@BLT.kr | www.BLT.kr
About the Author
Taekyun Chung, Partner Patent Attorney, serves as Head of Strategy at BLT Patent and Law Firm, where he advises startups on IP strategy, business model (BM) strategy, and go-to-market (GTM) strategy.
He graduated from Yonsei University with a degree in Electrical and Electronic Engineering and passed the 48th Korean Patent Attorney Examination in 2011. He currently works with startups across a wide range of industries, supporting their growth not only through intellectual property matters—including patents,
trademarks, and designs—but also through active involvement in their business strategy and development.
#SamsungSDI #AllSolidStateBattery #SolidStateBattery #AllSolidStateBatteries #BatteryTechnology #BatteryIndustry #NextGenerationBattery #AdvancedBattery #BatteryInnovation #EnergyStorage #BatteryManufacturing #BatteryProduction #BatteryManufacturingProcess #BatteryEquipment #RollPress #RollPressing #RollPressTechnology #WarmIsostaticPressing #WIP #SulfideSolidElectrolyte #SulfideBattery #AnodeFreeBattery #BatteryPatent #BatteryPatents #PatentAnalysis #PatentStrategy #PatentIntelligence #IPStrategy #IntellectualProperty #BLTDeepDive