Ohio State Innovation Foundation v. Akamai Tech Inc

I. Executive Summary and Procedural Information

• Parties & Counsel:
  • Plaintiff: Ohio State Innovation Foundation (Ohio)
  • Defendant: Akamai Technologies, Inc. (Delaware)
  • Plaintiff’s Counsel: K&L GATES LLP
• Case Identification: 1:19-cv-00484, E.D. Va., 04/19/2019
• Venue Allegations: Venue is based on Defendant's alleged regular and established place of business within the Eastern District of Virginia, where it is also alleged to commit acts of infringement.
• Core Dispute: Plaintiff alleges that Defendant’s content delivery platform and associated software development kit infringe a patent related to proactively pre-loading content onto mobile devices based on predicted user behavior to improve network efficiency.
• Technical Context: The technology addresses the challenge of network congestion by anticipating user data requests and shifting their delivery to off-peak times, a significant issue in managing traffic on mobile data networks.
• Key Procedural History: The complaint alleges a history of business discussions beginning in 2013 between the inventor's licensing entity and Defendant under a non-disclosure agreement. Following these discussions, Defendant allegedly declined a license, subsequently filed its own patent application on similar technology, and launched the accused products. The complaint also asserts that Plaintiff provided Defendant with pre-suit notice of infringement.

Case Timeline

Date	Event
2010-09-28	‘522 Patent Priority Date
2013-10-02	Non-Disclosure Agreement executed between Akamai and inventor's entity
2014-03-XX	Inventor's entity presents technology to Akamai
2014-12-XX	Akamai files patent application for similar technology
2016-02-XX	Akamai announces commercial release of its Predictive Content Delivery platform
2016-12-27	‘522 Patent Issue Date
2018-01-28	Plaintiff notifies Defendant of alleged infringement of the ‘522 Patent
2019-04-19	Complaint Filing Date

II. Technology and Patent(s)-in-Suit Analysis

• Patent Identification: U.S. Patent No. 9,531,522, "System and Method for Proactive Resource Allocation," issued December 27, 2016.

The Invention Explained

• Problem Addressed: The patent's background section identifies the "large disparity between the average and peak traffic demand" in wireless networks as a core problem that leads to the under-utilization of limited spectrum resources (’522 Patent, col. 1:42-44).
• The Patented Solution: The invention proposes a "proactive resource allocation framework" where the system exploits the predictability of user behavior to pre-emptively download content (’522 Patent, col. 2:15-17). Smart devices or network servers predict future requests, assign them deadlines, and schedule the data delivery in advance, often during network idle times. When the user initiates the actual request, the data is served from the device's local cache, reducing real-time network load (’522 Patent, col. 3:37-43).
• Technical Importance: This approach aims to smooth network traffic loads, thereby improving spectral efficiency and the user's Quality of Service without requiring additional physical network resources (’522 Patent, col. 1:21-31).

Key Claims at a Glance

• The complaint asserts independent claim 1 (’522 Patent, col. 13:41-col. 14:1; Compl. ¶8).
• The essential elements of claim 1 include:
  • A computerized method performed by a network server.
  • Collecting a plurality of prior requests for content from an individual user's mobile device.
  • Analyzing the prior requests with machine learning techniques to build a user profile.
  • Generating a list of anticipated requests with a deadline for each request.
  • For each anticipated request, downloading the corresponding data to the mobile device in advance of the deadline and storing it in memory.
  • Receiving an actual request for the content from the user.
  • Presenting the stored data from the device's memory in response to the actual request.
• The complaint does not explicitly reserve the right to assert dependent claims.

III. The Accused Instrumentality

Product Identification

• The complaint primarily accuses Akamai's Mobile Application Performance (“MAP”) SDK (’522 Patent, ¶¶ 6, 22). It also references Akamai's Predictive Content Delivery (“PCD”) platform as a related commercial release (’522 Patent, ¶21).

Functionality and Market Context

• The complaint alleges the MAP SDK provides the ability to "push content to the device before its requested" and "intelligently pre-loads content in the background" (’522 Patent, ¶22). This functionality is allegedly marketed as a way to "overcome times of high network congestion, or when the client may be entirely offline," allowing an application to "start up instantly, even when offline" (’522 Patent, ¶22).

IV. Analysis of Infringement Allegations

No probative visual evidence provided in complaint.

Claim Chart Summary

The complaint does not include its referenced claim chart exhibit. The following table summarizes the infringement theory based on the complaint's narrative allegations for representative claim 1.

Claim Element (from Independent Claim 1)	Alleged Infringing Functionality	Complaint Citation	Patent Citation
(a) collecting from a mobile user device a plurality of prior requests...	The MAP SDK platform is used to collect user data, which is then used to segment audiences based on behavior.	¶¶ 22, 28	col. 13:44-48
(b) analyzing said plurality of prior requests... with machine learning techniques to build a profile for said individual user;	The platform analyzes user data to "segment[] audiences and associat[e] key content with those segments," which allegedly constitutes building a user profile using the claimed techniques.	¶¶ 22, 28	col. 13:49-52
(c) generating a list of anticipated requests... said list comprising a deadline for each anticipated request;	The platform's ability to "push appropriate content during 'off hours'" is alleged to be based on a generated list of anticipated requests with an associated deadline or schedule.	¶¶ 22, 28	col. 13:53-55
(d) ... downloading to said mobile user device in advance of said deadline data ... storing said data in a memory...	The platform "intelligently pre-loads content in the background," which is stored on the device for later use.	¶¶ 22, 28	col. 13:56-61
(f) in response to said at least one actual request... presenting at said mobile user device data from said memory...	The pre-loaded content allows an application to "start up instantly, even when offline," which alleges that data is presented from local memory.	¶¶ 22, 28	col. 13:65-col. 14:1

Identified Points of Contention

• Scope Questions: A central question may be whether the accused system's function of "segmenting audiences" meets the claim's specific requirement of using "machine learning techniques" to build a profile for an "individual user." The complaint does not detail the specific techniques used by the MAP SDK.
• Technical Questions: The infringement analysis may turn on whether the MAP SDK's alleged scheduling of content delivery during "off hours" constitutes generating a "deadline for each anticipated request," as specified in claim 1. It raises the question of whether a general time window is equivalent to a per-request deadline as taught in the patent.

V. Key Claim Terms for Construction

The Term: "machine learning techniques"

• Context and Importance: This term's construction is critical because the infringement allegation hinges on whether Akamai's method for predicting content qualifies as "machine learning." Practitioners may focus on this term because the complaint's evidence, based on marketing materials, does not explicitly use this phrase to describe the accused product.
• Intrinsic Evidence for a Broader Interpretation: The term itself appears in the claim language, and the specification does not provide an explicit, limiting definition, which may support an argument for its plain and ordinary meaning (’522 Patent, col. 13:50-51).
• Intrinsic Evidence for a Narrower Interpretation: The detailed description emphasizes specific mathematical and statistical models, such as those based on large deviation theory and Poisson distributions, to achieve prediction (’522 Patent, col. 4:41-57). A party could argue the term should be construed in light of these specific disclosed embodiments rather than encompassing any predictive algorithm.

The Term: "deadline"

• Context and Importance: Claim 1 requires generating "a deadline for each anticipated request." The viability of the infringement claim may depend on whether the accused system's practice of pre-loading during "off hours" satisfies this limitation.
• Intrinsic Evidence for a Broader Interpretation: The specification describes the deadline as an "expiration time" for a request (’522 Patent, col. 4:1-2), a concept which could be argued to broadly cover any time-based constraint on pre-fetching.
• Intrinsic Evidence for a Narrower Interpretation: The patent repeatedly describes the deadline as a discrete value calculated for a specific request (e.g., 'Dq(n)=n+Tq(n)') and tied to a specific 'time slot' (’522 Patent, col. 4:2; Fig. 1). This could support a narrower construction requiring a distinct, calculated deadline for each item of content, as opposed to a general delivery window.

VI. Other Allegations

• Indirect Infringement: The complaint alleges that Akamai induces infringement by distributing the MAP SDK along with instructional information that encourages customers to use the platform in an infringing manner (Compl. ¶¶ 25, 29). It alleges Akamai had knowledge of the patent based on pre-suit business dealings and a specific notice of infringement sent in January 2018 (Compl. ¶¶ 18, 25, 29).
• Willful Infringement: Willfulness allegations are based on two main points: 1) alleged pre-suit knowledge of the technology and pending patents stemming from a 2013-2014 business relationship with the inventor's entity under an NDA, and 2) continued infringement after receiving explicit notice of the issued ‘522 Patent (Compl. ¶¶ 13-18, 25, 32).

VII. Analyst’s Conclusion: Key Questions for the Case

• A core issue will be one of definitional scope: can the claim terms "machine learning techniques" and "deadline for each anticipated request" be construed broadly enough to read on the accused system's described functionality of "segmenting audiences" and pre-loading content during general "off hours"?
• A key question will be one of technical evidence: what proof will be offered to demonstrate that the accused MAP SDK performs the specific steps of the claimed method, particularly the use of analytical techniques that rise to the level of "machine learning" and the generation of discrete, per-request "deadlines"?
• The litigation may also focus on the factual history between the parties. The allegations of pre-existing business discussions under an NDA, followed by Defendant’s development of a purportedly similar product, raise questions that could be central to the analysis of willfulness and potential damages.