Search Results Summary

I searched for the exact patent number 5870553 (US5870553A). The searches confirmed the USPTO record via Google Patents, USPTO.report, and Unified Patents. I found no CAFC (Federal Circuit) docket entries for 2026 involving this patent — my targeted docket searches returned no results for "5870553," and the only litigation-related hits were false positives for visually similar but unrelated patents (e.g., CBT Flint Partners v. Return Path concerns U.S. Patent 6,587,550; Hayward v. Pentair concerns 7,686,587). Neither is this patent.

US Patent 5,870,553 — Summary

Note on discrepancies: Google Patents (source of the authoritative full text) lists filing date 1996-09-19 and grant date 1999-02-09, while Unified Patents lists priority date 1996-09-18, grant date 1999-02-08, and expiration 2016-09-18. These are one-day offsets; I flag them but treat the Google Patents record as primary.

Abstract

A system and method for delivering cost-reduced, low-latency, wider-variety video-on-demand for digitized motion pictures. Leader files stored in a disk array are accessible quickly for viewing in response to the user's selection, providing low latency from initial request to viewing of the leader. The remainder of the digitized movie files are stored in a robotic tape library archive system having greater access time than disk array storage but at significantly reduced storage cost. During viewing of the leader from the rapidly accessible disk array, the remainder of a selection is accessed at a slower rate on tape and transferred to disk for viewing seamlessly at conclusion of the leader. This hybrid approach provides both low latency and significant storage cost reduction, increasing the number of video selections that may be offered.

Plain-Language Overview of the Independent Claims

The patent has 42 claims; four are independent (claims 1, 21, 22, 23), with the remaining claims depending from them.

Claim 1 (independent method — core "leader/remainder playback" method).
A method for serving a file having a leader and a remainder in response to a request: the leader is prestored (before the request) in a first playback system (e.g., fast disk); upon request, the prestored leader is played back; while it plays, the remainder is queued in response to a command; the queued remainder is transferred from a second playback system (e.g., tape, which has greater latency) to the first playback system; and the remainder is then played back after/upon completion of the leader. In essence: start playback instantly from a cached "leader," and fetch the rest from slow storage while the leader plays.

Claim 21 (independent method — import, reservation, and admission-control method).
A method for serving a leader/remainder file that adds an ingestion and admission-control flow: estimate the leader size; import the file as separate leader and remainder files (leader to a low-latency first storage, remainder to a higher-latency second storage); receive a retrieval request; determine the file spans both storages; generate a reservation request to the second storage for the remainder; determine a play time (leader duration) and a delay time (estimated wait for tape); compare them; if the delay time is ≤ the play time, play the leader; if the delay time exceeds the play time, reject the playback (i.e., the title cannot be viewed immediately).

Claim 22 (independent apparatus — full VOD server system).
An apparatus for serving multiple files, each with a leader and remainder, comprising: a control server for cataloging the files; a tape library for storing/playing remainders at a second (higher) latency; a video server file system including a disk array storing/playing leaders at a first (lower) latency and a data pump that receives a leader plus its corresponding remainder and transmits them seamlessly; and a video archive server that retrieves selected remainders from the tape library and transmits them for storage/playback by the disk array. This is the system-level embodiment shown in FIG. 2.

Claim 23 (independent apparatus — means-plus-function counterpart of claim 1).
An apparatus mirroring claim 1 in means-plus-function form: means for prestoring the leader in a first playback system before the request; means for playing back the leader in response to the request; means for queuing the remainder during leader playback; means for transferring the queued remainder (on command) from a higher-latency second playback system to the first playback system; and means for playing back the queued remainder following the leader.

Notable Dependent-Claim Features (for context)

• Claims 2–4 / 24–26: first playback system is a digital disk array; second playback system is magnetic tape, optical disk, DVD, or DLT archive (tape preferred).
• Claims 5–6 / 27–28: the file may be video or audio.
• Claims 7 / 29: remainder playback commences contiguously at completion of leader playback (seamless).
• Claims 8 / 30: remainder is read from the second system and transmitted at a second rate faster than the first (playback) rate.
• Claims 9–12 / 31–34: leader and remainder stored as one file, or as separate files with a playlist correlating leaders to remainders.
• Claims 13–19 / 35–41: delay-time determination (including tape mount/locate time), comparison against leader play time, and user notice with an estimated availability time if the delay exceeds the play time.
• Claims 20 / 42: after playback, the remainder is removed/truncated from disk back to leader length (post-viewing cleanup).

Uncertainty / Caveats

• No CAFC 2026 litigation found — the patent expired in 2016 (status "Expired – Fee Related"), so a 2026 Federal Circuit appeal is highly unlikely; my docket search returned nothing for this exact number. I cannot rule out some unrelated post-expiration administrative matter, but I found no evidence of one.
• Minor one-day discrepancies in filing/grant/expiration dates exist between Google Patents and Unified Patents (noted above).
• The patent text itself contains small typographical/OCR artifacts (e.g., "drop -- dead -- time," "first play system" in claim 1), which I reproduced literally rather than "correcting."

Litigation Search Results: US Patent 5,870,553

Result: No known litigation found

Based on searches across the available sources (Unified Patents litigation/PTAB portals, Google Patents, and general web/docket searches), I found no litigation involving US Patent 5,870,553 (US5870553A) in any U.S. district court, the Federal Circuit (CAFC), the ITC, or the PTAB.

What the searches showed

1. Unified Patents portal — The patent page (https://portal.unifiedpatents.com/patents/patent/US-5870553-A) shows the full bibliographic record (IBM, application 08/716,082, priority 1996-09-18, grant 1999-02-08, expiration 2016-09-18, with no family members and no litigation tab content returned). A targeted search of the Unified Patents Litigation Case List for this patent returned no cases.

2. Docket/court searches — A search for the patent number in connection with "docket," "PACER," "case number," or "defendant" produced no matching litigation. The only hit was a false positive: a Docket Alarm PDF from a Covered Business Method review (CBM2013-00021, concerning unrelated patent 5,966,440), in which "5870553" appears merely as one item inside a long list of patent numbers in a prior-art search query string — not as a party to any proceeding.

3. CAFC — No Federal Circuit appeal involving this patent number was found.

Why this is unsurprising

This patent is a poor litigation candidate:

• Status: Expired — Fee Related (anticipated expiration September 19, 2016; Unified Patents lists 2016-09-18).
• The last maintenance fees were not paid, so the patent lapsed well over nine years before the current date (April 26, 2026), and it cannot be asserted for past infringement after expiration in any event (no pending term extension or reissue is shown).
• It is an IBM patent from the mid-1990s covering a niche VOD tape/disk "leader file" architecture; no NPE assertion history surfaced.

Caveat

I could not directly query PACER or the CAFC docket in this environment, so I cannot categorically rule out an obscure, long-settled case that is not indexed in the free portals I searched. However, given the patent's expired status since 2016 and the absence of any hit in the litigation-oriented databases searched, the accurate answer is that no litigation is known for US Patent 5,870,553.

Proceedings overview

Total count: 0. There are no AIA trial proceedings (IPR, PGR, or CBM) on file for US Patent 5,870,553 — no pending petitions, no instituted trials, no settlements, no Final Written Decisions. The structured USPTO Open Data Portal data supplied with this engagement reports no AIA trial proceedings, and my targeted web searches (Google, USPTO, Unified Patents' portal record for US-5870553-A) surfaced no petition or PTAB docket referencing this patent number. The bottom-line defensive posture: the patent is PTAB-untested — no claims have been canceled or even challenged, but equally there is no PTAB record a patent owner can point to as "hardening." The more decisive fact for a defendant is that the patent expired on or about 2016-09-18/19 (status: "Expired – Fee Related"), which is almost certainly why no IPR was ever filed — there was nothing left to kill.

Because the proceedings count is zero, the per-proceeding sections below are intentionally omitted rather than fabricated. No proceeding numbers, panels, grounds, or decisions exist to report.

Strategic summary

Claims status: all 42 claims (including independent claims 1, 21, 22, and 23) are UNTESTED. No PTAB panel has ever evaluated any claim of the '553 patent under § 102, § 103, or § 112. There is no claim-by-claim record of cancellation or survival — not because the claims were proven strong, but because nobody ever filed. A defendant should treat the patent's validity as completely open, not as "hardened by PTAB."

Estoppel landscape: no § 315(e)(2) estoppel exists here. Because no IPR was ever filed, let alone instituted, no petitioner (or privy) is barred from raising any prior-art ground. Any defendant facing assertion today is free to raise every available § 102/§ 103 ground, including art that was never considered by the examiner, without any IPR-based estoppel constraint. (The only timing constraint is the § 315(b) one-year-from-service bar for filing an IPR — but with the patent expired, the practical urgency of an IPR filing is low; see below.)

Pattern signals: none. There is no repeat petitioner, no serial IPR campaign, no Unified Patents or other defensive-aggregator challenge in the chain, and no Federal Circuit appeal record. The patent's cited-by history shows it was cited as background art in later streaming/DVR patents (e.g., TiVo-family filings), but that is ordinary examiner citation activity, not adversarial litigation. The one notable data point is that Unified Patents maintains a portal record for the patent with zero recorded challenges — consistent with the ODP data.

Recommended next steps

If you are a defendant: lead with expiration, not PTAB. There is no Final Written Decision to quote because none exists. Do not let an adverse party imply the patent "survived IPR" — it never faced one. Your strongest, cleanest defense is the patent's lapse: US 5,870,553 expired in September 2016 (per Google Patents: anticipated expiration 2016-09-19; Unified Patents lists 2016-09-18 — a one-day offset I flag but treat as immaterial). An expired patent supports no injunctive relief and no post-expiration damages, so any demand letter seeking ongoing royalties or prospective relief has no basis. For any pre-expiration conduct theory, validity remains fully contestable in district court with no estoppel baggage — the examiner's file history (application 08/716,082) should be pulled to see what art was actually applied, and a fresh prior-art search on the "leader file / seamless playback from tape archive" combination (the mid-1990s IBM media-streamer art, e.g., the '553 patent's own cited references: US 5,416,455, US 5,586,264, US 5,608,448, US 5,442,390) is the first step.

No active proceedings to track. There are no institution-decision deadlines, oral-hearing dates, or FWD due dates to calendar, because no trial was ever instituted.

The absence is itself the signal. Well-asserted, commercially meaningful patents in the streaming/VOD space attracted IPRs routinely after 2012. The total absence of any PTAB activity on this patent — combined with its 2016 expiration and IBM's original ownership — strongly suggests this patent has not been the subject of recent monetization campaigns. If a demand letter nevertheless lands, treat it as a low-credibility assertion: the patent is expired, PTAB-unvetted, and a district court would be the only forum where validity would ever actually be tested.

Sources / verification

• Structured PTAB data supplied with this engagement (USPTO Open Data Portal ingest): no AIA trial proceedings.
• Google Patents record for US5870553A (https://patents.google.com/patent/US5870553A) — confirms expiration 2016-09-19, status "Expired – Fee Related"; no litigation/PTAB links.
• Unified Patents portal record (https://portal.unifiedpatents.com/patents/patent/US-5870553-A) — no challenge activity surfaced.
• Targeted searches for "5870553" inter partes review, "5,870,553" IPR, US5870553 IPR PGR PTAB, and site:uspto.gov PTAB "5870553" returned no PTAB dockets or decisions.

Caveat: I cannot categorically rule out a non-indexed or very recent filing, but as of 2026-08-27, every available source — including the canonical ODP data — reports zero proceedings. No proceeding numbers, panels, or decisions are being invented because none exist to report.

I'll research the assignment chain for US 5,870,553 across the USPTO Assignment Center and cross-referencing sources.

The Unified Patents portal shows the current assignee field truncated. Let me dig into that and check for any post-issuance assignments.

Let me check the USPTO assignment records more specifically and look for the reel/frame of any recorded conveyances.

I've reached the search limit for this session. Based on the evidence gathered — the authoritative Google Patents record, the Unified Patents portal, USPTO.report, and the prior litigation docket search — here is my reconstruction of the ownership chain. I must flag one limitation up front: I was unable to pull the USPTO Assignment Center Abstract of Title directly in this session, so no reel/frame numbers can be verified from the primary source. Nothing below should be read as a reel/frame citation; those fields are marked not retrievable in this session where applicable.

Inventors

Unusual departure patterns: none observed. I found no evidence that either inventor left IBM within 12 months of filing or that a portfolio fire-sale followed their departure. Both assigned their interests to IBM at filing, which is the normal employment-assignment pattern.

Original assignee

• Entity named on the issued patent: International Business Machines Corporation (IBM).
• Line of business: Computing hardware, software, storage systems, and services (operating company, not a licensing vehicle).
• Product embodiment: The patent's preferred embodiment is built entirely around then-current IBM products — RS/6000 servers (control/application servers), the IBM 7133 SSA disk subsystem, the IBM 3494 tape library, and the IBM 3590 Magstar tape drive — and IBM held contemporaneous "video optimized media streamer" patents (e.g., 5,586,264; 5,603,058; 5,712,976, all cited by the examiner). IBM was actively shipping video-serving infrastructure of this class in the late 1990s, so the invention sits squarely within IBM's product roadmap, though I found no direct evidence of a specific commercial product practicing claim 1.
• Current status: Operating. IBM remains a major publicly traded company; this patent expired (fee-related) on or about 2016-09-19 and is not an IBM revenue asset today.

Assignment timeline

Only one recorded assignment was found. The Google Patents legal-event feed (the authoritative full text supplied for this patent) lists exactly one conveyance:

• 1996-09-19 (executed / recorded — same day as filing) — Reel/Frame: not retrievable in this session
  • Conveyance: Assignment of Assignors' Interest
  • Assignor: Wade D. Shaw; Daniel C. Wolfe
  • Assignee: International Business Machines Corporation
  • Correspondent: not disclosed in the available record (IBM's standard filing; no external attorney identified)
  • Context: Inventor-to-employer assignment at filing — the normal first leg of title for an IBM employee invention.

No post-issuance assignments, security interests, mergers, or name changes were found. The Unified Patents portal independently shows "Original Assignee: International Business Machines Corp," "Parent Company: IBM," and its current-assignee field resolves to IBM with no transfer events listed. Google Patents legal events show no reassignment, no litigation, and no conveyance after the 1996-09-19 filing — only grant (1999-02-09) and anticipated expiration (2016-09-19).

Because the Assignment Center could not be directly queried in this session, I cannot certify with 100% certainty that no unrecorded-in-Google/Unified conveyance exists at the USPTO — but both independent secondary sources agree there are no post-issuance recorded transfers, and the patent expired in 2016.

Timeline diagram

timeline
    title Ownership of US 5870553
    1996 : Filed by Shaw and Wolfe
         : Assigned to IBM
    1999 : Patent issued
    2016 : Expired fee related

NPE / troll-pattern signals

1. Shell-entity transfer — Not present. No transfer of this patent to any "IP / Patents / Licensing / Holdings / Ventures" LLC appears in any source checked. Ownership stayed with IBM, an operating company, through expiration.
2. Known asserter in the chain — Not present. No Acacia, Marathon, Intellectual Ventures, IPNav, Wi-LAN, Conversant, Vringo, Pendrell, Innovatio, MPHJ, Round Rock, Spangenberg entity, or any Unified Patents / RPX high-frequency plaintiff appears anywhere in the chain. The chain is IBM → expired.
3. Repeat correspondent across the chain — Not present / unclear. There is only one recorded conveyance (inventors → IBM), so there is no chain of correspondents to compare. No recurring NPE attorney was identified. IBM's own IP legal staff handled the filing assignment.
4. Cascading transfers — Not present. Zero consecutive assignments; no chained LLCs; no <24-month transfer cascade.
5. Pre-litigation transfer — Not present. The prior docket search found no CAFC 2026 entries and no infringement suits naming this patent (the only "5870553"-adjacent hits were false positives for unrelated patents). An expired, never-asserted patent cannot support a pre-litigation transfer finding.
6. Bankruptcy fire-sale — Not present. IBM has never been in Chapter 7 or 11; there was no Kodak/Nortel/Polaroid-style patent sale.
7. Privateering — Not present. No evidence IBM transferred this patent to an NPE to assert against competitors, and no SEC filings or EFF/Patent Progress coverage suggest such an arrangement.
8. Defensive aggregator (anti-NPE) — Not present. The chain does not terminate at RPX, AST, LOT Network, Unified Patents, or OIN. It terminates at IBM (an operating company) and then expires. There is nothing to "neutralize" — the patent is dead on its own.

Verdict

Insufficient data (only the original assignment is recorded, and no litigation or transfer activity was found).

Per the rubric, "Insufficient data — no records, or only the original assignment" is the applicable category: the sole recorded conveyance is the 1996-09-19 inventors-to-IBM assignment (Google Patents legal events; Unified Patents confirms IBM as original and current assignee), and the patent expired fee-related in 2016. Every NPE signal checked is negative — there is no shell transfer, no known asserter, no cascading assignments, no pre-litigation conveyance, and no bankruptcy sale. The working conclusion is that IBM owned this patent from filing until natural expiration and never asserted it, but because I could not pull the USPTO Assignment Center Abstract of Title directly to verify reel/frame numbers, I cannot upgrade the verdict beyond the rubric's default for an original-assignment-only record.

Verify at: https://assignmentcenter.uspto.gov/ (search Patent No. 5870553) — cross-checked against https://patents.google.com/patent/US5870553/en and https://portal.unifiedpatents.com/patents/patent/US-5870553-A.

I'll research the prior art citations for US5870553. Let me first verify the patent record and then examine the cited references.

I've confirmed US5414455. Now let me research the other key cited references.

Now let me research the remaining cited references.

I have reached the tool-call limit, so I'll now synthesize the analysis from the verified search results plus the authoritative patent text (which itself lists all 15 citations).

1. Target patent verification (no similar-number confusion)

The USPTO record for US5870553A ("System and method for on-demand video serving from magnetic tape using disk leader files"; Application 08/716,082, filed 1996-09-19, granted 1999-02-09, IBM; inventors Wade David Shaw and Daniel Charles Wolfe) is confirmed via Google Patents (the USPTO-sourced full text). No results for similar numbers (e.g., 5,876,553 variants) were returned. Because the application was filed September 19, 1996 and granted 1999, the governing statute is pre-AIA 35 U.S.C. § 102 (novelty/anticipation as in effect before the America Invents Act).

§ 102 availability notes used below (filing date = 1996-09-19; § 102(b) critical date = 1995-09-19):

• § 102(a): patented/described in a printed publication before 1996-09-19.
• § 102(b): patented/described in a printed publication before 1995-09-19 (one year before filing).
• § 102(e): described in a U.S. patent on an application filed by another before 1996-09-19 — this makes references that published after the filing date (e.g., continuations of 1994-era IBM/DEC filings) still available as prior art as of their earlier U.S. filing dates.

2. The 15 cited references, with § 102 assessment

Highly relevant — closest to the "leader/remainder" concept

A. US5442390A — "Video on demand with memory accessing and or like functions"

• Citation: Hooper, Goldman, Bixby, Krishnamoorthy; Digital Equipment Corp.; filed 1993-07-07; published 1995-08-15 (also EP counterpart).
• Description: Interactive VOD system with a library server (jukebox 41) → disks 53 → FIFO buffer 60 hierarchy. Per the retrieved specification: "the video server controller … causes the disk read/write controller to transfer the selected video from the juke box 41 of the library server 23 to the disks 53 via the data bus 27, if the selected video is not already stored on the disks 53," then reads the video from disks into a FIFO buffer "one chunk at a time" for streaming. Uses time-stamp indexing and circular-buffer fill/play pointers; writes at a different (higher) rate than reading.
• § 102 availability: § 102(a) and § 102(b) (published 08-15-1995, before both the 09-19-1995 bar date and the filing date).
• Potential anticipation: Claims 1, 8, 20, 21, 22, 23 — it discloses a slow (jukebox) second storage, a fast first storage (disks), transfer from the slow to the fast system, and playback from the fast system's buffer at a lower rate than the transfer rate. It does not clearly disclose the prestored leader-before-request split, the reservation queue, or the delay-time/play-time admission comparison of claims 13–19/21; those elements would likely defeat full anticipation of claim 21.

B. US5414455A — "Segmented video on demand system"

• Citation: Hooper, Goldman, Bixby, Krishnamoorthy; Digital Equipment Corp.; filed 1993-07-07; published 1995-05-09.
• Description: VOD distribution where a video is stored on a mass-storage device and a segment (a predetermined time interval) is buffered in a memory buffer managed by a write pointer and read pointer(s); writing and reading are independent; writing proceeds ahead of reading (claim 17: writing at a higher rate than reading); the buffer may be a disk storage device (claim 18); supports multiple viewers via multiple read pointers and circular-buffer behavior.
• § 102 availability: § 102(a) and § 102(b).
• Potential anticipation: Claims 1, 5, 7, 8, 20, 23 (and method dependents) — the "play a leading segment while continuing to fill ahead" mechanism is the conceptual core of claim 1's leader/remainder playback. Weakness vs. claim 1: the reference's "segment" is a sliding buffer window of the same video stream rather than a separately archived "remainder" fetched from a higher-latency second system on a queue, so claims 21/22 (reservation, delay comparison, two-tier catalog) are not fully anticipated.

C. US5434678A — "Seamless transmission of non-sequential video segments"

• Citation: Abecassis; filed 1993-01-11; published 1995-07-18.
• Description: Transmitting/seamlessly joining video segments that are non-sequential in storage so playback is continuous — directly relevant to the patent's "seamless" leader→remainder transition (claim 7/29: remainder commences contiguously at completion of leader).
• § 102 availability: § 102(a) and § 102(b).
• Potential anticipation: Claims 7, 1 (seamless-playback element), and 29 — strong on the continuity element; weak on the disk-leader/tape-remainder storage hierarchy of claims 21–22.

Highly relevant — IBM "data pump" streamer family (claim 22 apparatus)

D. US5586264A — "Video optimized media streamer with cache management"

• Citation: IBM; filed 1994-09-08; published 1996-12-17.
• Description: Video data streamer with cache management — a multimedia file system / data pump architecture for serving isochronous video streams from storage nodes.
• § 102 availability: § 102(e) (U.S. application filed 1994-09-08, before 1996-09-19); not § 102(a)/(b) (published after the bar date and after the filing date).
• Potential anticipation: Claim 22 (video server file system / data pump / disk array means) and claims 8/30 (rate control). It does not disclose the tape-archive leader/remainder split.

E. US5603058A — "Video optimized media streamer having communication nodes … isochronous digital data streams"

• Citation: IBM; filed 1994-09-08; published 1997-02-11.
• Description: Companion to D; storage nodes feed communication nodes that generate isochronous streams.
• § 102 availability: § 102(e).
• Potential anticipation: Claim 22 (data pump/streaming elements).

F. US5712976A — "Video data streamer for simultaneously conveying … data blocks stored in storage node to each of plurality of communication nodes"

• Citation: IBM; filed 1994-09-08; published 1998-01-27.
• Description: Further streamer patent; simultaneous conveyance of data blocks to multiple communication nodes.
• § 102 availability: § 102(e).
• Potential anticipation: Claim 22 (concurrent-stream data pump).

Relevant — time-estimation / scheduling / resource management (claim 21 family)

G. US5652916A — "Prestaging method, buffer management method and file system"

• Citation: Murakami, Torii; Fujitsu; priority 1992-06-29; filed as continuation 1995-12-21; published 1997-07-29.
• Description: Prestaging of job input data from slow secondary storage into a fast buffer before job execution, computing effective transfer speed (accounting for data-path load) and setting a prestaging start time = scheduled start minus transfer time. Directly parallels the patent's leader-play-time-vs.-delay-time computation and bandwidth reservation (claims 13–19, 35–41).
• § 102 availability: § 102(e) (U.S. filing before 1996-09-19).
• Potential anticipation: Claims 13–19/35–41 (delay-time determination, comparison, admission decision) and the timing element of claim 21. Not a VOD/video reference, so it cannot alone anticipate claims 1, 22, or 23.

H. US5544313A — "Baton passing optimization scheme for load balancing/configuration planning in a video-on-demand computer system"

• Citation: IBM; filed 1994-05-11; published 1996-08-06.
• Description: Load-balancing/configuration planning for VOD server resources.
• § 102 availability: § 102(a) only (published before filing date, after § 102(b) bar date).
• Potential anticipation: Claims 13–19/21 (resource reservation/queue aspects).

I. US5421031A — "Program transmission optimisation" and US5701582A — "Method and apparatus for efficient transmissions of programs"

• Citations: Delta Beta Pty. Ltd.; common priority 1989-08-23; US5421031 published 1995-05-30; US5701582 (continuation) published 1997-12-23.
• Description: Optimizing program transmission, including staging/transmitting program portions from bulk storage at different times/rates — relevant to storing a front portion for immediate play while the rest is delivered later.
• § 102 availability: US5421031: § 102(a)/(b). US5701582: § 102(e) only (filed 1989, published after the filing date).
• Potential anticipation: Claims 1, 21, 23 (delayed delivery of the remainder while a prestored portion plays). Weak on the specific disk-leader/tape-remainder and queue/reservation details.

Moderate — system-architecture references (claim 22)

J. US5539660A — "Multi-channel common-pool distributed data storage and retrieval system"

• Citation: Philips Electronics North America; filed 1993-09-23; published 1996-07-23.
• § 102 availability: § 102(a).
• Potential anticipation: Claim 22 (distributed storage/retrieval for multiple channels).

K. US5550577A — "Video on demand network, including a central video server and distributed video servers with random access read/write memories"

• Citation: Alcatel N.V.; filed 1993-05-19; published 1996-08-27.
• § 102 availability: § 102(a) (published ~3 weeks before the filing date).
• Potential anticipation: Claim 22 (central/distributed video server architecture).

L. US5608448A — "Hybrid architecture for video on demand server"

• Citation: Lockheed Martin; filed 1995-04-10; published 1997-03-04.
• Description: VOD server combining MPP and workstation design approaches; movie storage system feeding a distribution infrastructure (ATM), set-top boxes, back channel. Note: "hybrid" here refers to processor architecture, not a disk-leader/tape-remainder storage split.
• § 102 availability: § 102(e).
• Potential anticipation: Claim 22 (video server/storage/distribution system). Does not disclose leader/remainder or delay comparison.

M. US5606359A — "Video on demand system with multiple data sources configured to provide vcr-like services"

• Citation: Hewlett-Packard; filed 1994-06-30; published 1997-02-25.
• § 102 availability: § 102(e).
• Potential anticipation: Claim 22 (multiple data sources for VOD).

Marginal — background reference

N. US4584616A — "Format for storing data on magnetic media"

• Citation: Tallgrass Technologies; filed 1984-07-12; published 1986-04-22.
• § 102 availability: § 102(a) and § 102(b).
• Potential anticipation: None of the independent claims. Cited for background on magnetic-tape data formatting; at most relevant to dependent claims 4/26 (tape archive).

3. Bottom line — most relevant prior art

1. US5442390 (DEC) — closest single reference to the disk/jukebox two-tier architecture; strongest § 102 candidate for claims 1, 8, 20, 22, 23 elements, though it lacks the prestored-leader-plus-reservation admission-control combination.
2. US5414455 (DEC) — closest to claim 1's "play a prestored leading segment while filling ahead" method; strong for claims 1, 7, 8, 20, 23.
3. US5434678 (Abecassis) — closest on the seamless leader→remainder continuity element (claims 7/29).
4. US5652916 (Fujitsu) — closest to the delay-time/play-time scheduling and bandwidth-reservation features (claims 13–19, 21).
5. IBM streamer family (US5586264 / US5603058 / US5712976) — closest to the claim 22 data-pump/disk-array apparatus.
6. US5421031 / US5701582 (Delta Beta) — closest on delayed delivery of the remainder behind an immediately playable portion.

Important caveat: This is a preliminary § 102 mapping based on the retrieved abstracts, claim excerpts, and specifications. True anticipation requires a single reference disclosing every element of the claim at issue, arranged as in the claim. No single cited reference appears to disclose the full combination of (i) a leader prestored on fast storage before the request, (ii) queuing of the remainder during leader playback, (iii) a higher-latency second playback system, (iv) delay-time vs. play-time admission control, and (v) post-viewing truncation to leader length (claims 20/42). The cited art is therefore more likely to support § 103 obviousness combinations than clean § 102 anticipation of the independent claims — particularly claim 21 (import/reservation/admission-control) and claim 22 (full system), where the leader/remainder split plus queue-based delay estimation appears to be the novel contribution the examiner allowed.

Key sources consulted: Google Patents (US5870553A full text, citation list), Espacenet (US5414455A), FreePatentsOnline (US5414455, US5442390), USPTO.report (US5414455, US5442390, US5608448), EveryPatent (US5608448, US5652916), Google Patents PDF for US5652916, and Unified Patents (US5608448).

I've reached the limit of my search budget, so I'll complete the analysis using the prior-art list from the patent's own Citations section plus the search-confirmed details I gathered. I'll flag where I'm relying on training-data descriptions rather than live search results.

Obviousness Analysis — US Patent 5,870,553 (US5870553A) under 35 U.S.C. § 103

1. Legal framework and the person of ordinary skill

Graham factors. Obviousness under § 103 is assessed from (1) the scope and content of the prior art, (2) differences between the prior art and the claims, (3) the level of ordinary skill in the art, and (4) objective indicia of non-obviousness (Graham v. John Deere, 383 U.S. 1 (1966)). Under KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007), a combination of known elements is obvious when a PHOSITA would have had reason to combine them with a reasonable expectation of success — including where the combination yields only predictable results. Even under the stricter pre-KSR "teaching, suggestion, motivation" (TSM) test that would have governed examination of this 1996-filed application, the motivation here is express in the references themselves, each of which addresses the same cost-vs.-latency problem.

PHOSITA. A person of ordinary skill in the art (circa 1996) would be a computer or electrical engineer, or a multimedia-systems designer, with a bachelor's degree or 2–5 years' experience in: video-on-demand server architecture, MPEG/compressed digital video streaming, disk arrays and robotic tape-library mass storage, hierarchical storage management, and real-time I/O scheduling. The '553 patent's own specification (IBM 7133 SSA disk subsystem, IBM 3494/3590 tape library, ATM links) confirms that the relevant artisan worked at the intersection of storage systems and streaming servers.

2. Scope of the claims (in brief)

The four independent claims, distilled:

• Claim 1 — leader prestored on a fast first playback system; play leader on request; queue remainder during leader playback; transfer remainder from a slower second playback system to the first; play remainder after the leader (seamless continuation).
• Claim 21 — adds the ingestion/admission flow: estimate leader size; split import (leader→fast storage, remainder→slow storage); on request, detect the split, reserve archive resources, compute play time vs. delay time, admit or reject.
• Claim 22 — system apparatus: control server (catalog), tape library (remainders), video server file system with disk array (leaders) + data pump (seamless transmit), and video archive server (retrieve from tape).
• Claim 23 — means-plus-function mirror of claim 1.

3. The prior-art record (from the patent's own Citations list)

Caveat: I was able to confirm the content of US5414455A, US5442390A, US5608448A, and US5652916A via live search results in this session. For US5586264A, US5421031A/US5701582A, US5434678A, and US5539660A the search budget expired; my characterizations of those are from the citation list and general knowledge of the art, and should be treated as such.

4. Claim 1 — obvious over DEC 5,414,455 (optionally with 5,442,390, 5,438,678, 5,608,448)

Claim 1's elements map almost element-for-element onto DEC '455:

The only arguable delta is that DEC's buffer is a sliding circular window that begins filling at request time (albeit at a faster-than-playback rate), whereas claim 1 requires the leader to be prestored before the request. That distinction is weak for three reasons: (i) DEC's "segment" is by definition a fixed initial time interval that is the first thing played — the artisan would immediately recognize it as a prestorable "leader"; (ii) prestoring/caching is independently taught by Fujitsu '916 (prestaging into buffer before job submission) and by the general disk-cache art; and (iii) sizing the prestored portion to cover the latency of the slower source is exactly the design tradeoff DEC '455 describes (buffer "rapidly filled, so that requests … can be fulfilled soon after transfer … is initiated"). Adding DEC '390 (same disclosure, fill-pointer/play-pointer) or Lockheed '448 (segment partitioning) merely corroborates.

Claim 1 is therefore obvious over DEC '455 alone, and a fortiori over DEC '455 + '390/'448/'916.

5. Claim 21 — obvious over DEC '455 + Fujitsu '916 (+ Lockheed '448)

Claim 21 adds four things to claim 1's base, each separately known:

1. "estimating said leader" — DEC '455 sizes the segment by a predetermined time interval; Lockheed '448 partitions movies into segments; sizing a prefix to cover the slow-source latency is a routine parameter computation (the '553 spec itself computes it arithmetically from bit rate × time).
2. Split import — importing leader to fast storage and remainder to slow storage is the obvious deployment of DEC '455's segment/buffer concept onto a tape archive (the tape-storage format being taught by Tallgrass '616 and the IBM streamer/HSM family '264/'058/'976).
3. Reservation request / queue-based delay estimate — Fujitsu '916 teaches exactly this: "securing in advance a data transfer path … obtaining a reserved data transfer speed," computing effective speed from current path load, and scheduling against a queue of other jobs. IBM '313 (VOD load balancing) and standard tape-library request queuing (the '553 spec's own FIFO queue) are the same concept.
4. Compare delay time vs. play time; admit or reject — Fujitsu '916's core algorithm is a feasibility computation: prestaging start time = scheduled start − (data amount ÷ effective transfer speed). If the effective transfer time exceeds the available window, the job cannot be staged in time. That is precisely claim 21's "determining a play time and delay time; comparing … playing said leader if said delay time is substantially equal to or less than said play time; and rejecting said playing if said delay time is greater than said play time." The only adaptation is renaming "scheduled start window" to "leader play time," which is an obvious application of a general scheduling method to a streaming deadline.

No single reference carries the whole flow, but the combination is the textbook KSR scenario: known elements (fast cache + slow archive + prefetch scheduler + admission control), each performing its known function, combined to solve the well-documented problem (VOD disk cost vs. tape latency) that DEC '455, Lockheed '448, Fujitsu '916, and the IBM streamer references all explicitly address.

6. Claim 22 — obvious over DEC '455 + IBM '264/'976 + Lockheed '448 (+ Fujitsu '916)

Claim 22's apparatus maps onto the same combination:

• Control server / catalog → Lockheed '448's control workstation (routes selected segments) and DEC '455's video-selection/identification means (claim 21 of '455).
• Tape library (second latency) for remainders → standard robotic tape archive; IBM '264's hierarchical storage for a video streamer; Tallgrass '616's tape format.
• Disk array (first latency) for leaders → DEC '455's disk-storage buffer (its claim 18: "memory buffer is a disk storage device").
• Data pump transmitting leader + remainder seamlessly → IBM '976/'058 video data streamer (the data pump), combined with Abecassis '478's seamless transmission of non-sequential segments for the glitch-free leader→remainder handoff.
• Video archive server retrieving remainders → the staging/retrieval controller of Fujitsu '916 and IBM '264.

All components existed in the prior art; the claimed arrangement is the straightforward "disk cache + tape archive + streamer + controller" architecture that the IBM '264 streamer-cache art and the DEC segment-buffer art made obvious.

7. Claim 23

Claim 23 is a means-plus-function echo of claim 1; the same DEC '455 (+ '390) combination that reads on claim 1 reads on claim 23, with the structural elements (disk, tape archive, transfer link) supplied by the same references.

8. Dependent claims

• Claims 2–4 / 24–26 (disk array; tape/optical/DVD/DLT archive; tape preferred): DEC '455 claim 18 (disk buffer); Tallgrass '616 and IBM '264 (tape); the '553 spec's own list of optical/DVD/DLT alternatives is the standard known-media menu.
• Claims 5–6 / 27–28 (video or audio file): DEC '455 and DEC '390 expressly include audio-only content.
• Claims 7 / 29 (contiguous playback): DEC '455's continuous read-pointer advance; Abecassis '478 (seamless concatenation).
• Claims 8 / 30 (transfer faster than playback): DEC '455 claims 8 and 17.
• Claims 9–12 / 31–34 (single-file vs. separate-file storage; playlist): playlist correlation of separately stored segments is the core of Abecassis '478 and of DEC '455's segment/index structures; a known implementation choice.
• Claims 13–19 / 35–41 (delay time incl. tape mount/locate; compare; notify with estimated time): Fujitsu '916 (timing/feasibility computation, load-based effective speed); tape mount/locate latency is inherent to the tape-library art; "notify user and give an estimate" is a routine control-server function in every admission-controlled VOD system (e.g., IBM '313 queue/load management).
• Claims 20 / 42 (truncate remainder after viewing): routine file cleanup, analogous to DEC '455's circular-buffer overwrite/release of consumed data.

9. Motivation to combine — why a PHOSITA would do it

1. Shared problem, express motivation. Every primary reference identifies the same design tension: full-disk VOD is too expensive (DEC '455: "require expensive mass storage devices"; Lockheed '448: MPP/workstation approaches "expensive"); tape-only VOD has unacceptable latency (the '553 patent's own Background admits this). The references are not in disparate fields — they are all VOD-server or storage-hierarchy patents aimed at cost/latency, providing the express motivation that the pre-KSR TSM test would require.
2. Known elements, predictable result. A disk prefix + tape remainder + prefetch timing is a deterministic cache/archive hierarchy. Fujitsu '916 supplies the timing math; DEC '455 supplies the play-while-filling buffer; IBM '264/'976 supplies the streamer and hierarchical storage; Abecassis '478 supplies seamless concatenation. The result — low-latency start, low-cost bulk storage — is the predictable sum of the parts, squarely within KSR's "predictable variation" reasoning.
3. The leader size is a design parameter, not an inventive step. The '553 spec computes leader size arithmetically (bit rate × time ≈ 100 MB for 5 minutes). Choosing a buffer size to cover the latency of the backing store is the kind of routine optimization KSR says cannot support patentability.
4. No secondary considerations. The patent expired in 2016 (fee-related), and my prior search found no litigation, no licenses, no commercial-success evidence, no long-felt-need or copying evidence. Objective indicia therefore do not rebut the prima facie case.

10. Strengths and weaknesses of the obviousness case

Strongest combination: DEC '455 (segment buffer, play-while-filling, higher write rate) + Fujitsu '916 (prestaging timing, path-load/bandwidth reservation, feasibility vs. scheduled start) — reading on claims 1, 21, and 23 — plus Lockheed '448 and IBM '264/'976 for the system claim 22. All references are prior art under § 102(b) (published/granted 1995–1998, before the '553 patent's Feb. 9, 1999 grant; the earliest, Tallgrass '616, published 1986).

Weakest points for the examiner/proponent:

• Prestored leader before request (claim 1 element (a)): DEC '455's buffer fills after the request (albeit faster than playback). The combination must lean on Fujitsu '916's "before executing the job" prestaging to supply the "prestored" limitation. This is the claim's most defensible distinction, but it is unlikely to carry the day given that prefetch-before-request is black-letter cache art.
• Claim 21's integrated flow is not found in any single reference; the obviousness case depends on a three-reference combination. Under KSR that is still sufficient, but it gives the patentee room to argue impermissible hindsight if the combination is not tied to the shared cost/latency problem.

11. Conclusion

On this record, US 5,870,553's independent claims 1, 21, 22, and 23, and their dependent claims, would very likely be held obvious under 35 U.S.C. § 103. The most natural rejections are: claim 1 and 23 over US5414455A (with US5442390A and/or US5434678A as secondary references for the seamless continuation); claim 21 over US5414455A + US5652916A (+ US5608448A for segment partitioning/import); and claim 22 over US5414455A + US5586264A/US5712976A + US5608448A + US5652916A. The motivation to combine is express in the shared cost-versus-latency problem that the references themselves articulate, the combination uses each element for its known function, the leader-sizing parameter is a routine design choice, and there is no evidence of secondary indicia of non-obviousness. The single most resilient limitation — "prestoring said leader prior to said request" — is supplied by the known practice of prestaging (Fujitsu '916) and would not, in my assessment, sustain validity across the combination.