DCT
1:18-cv-00318
IPA Tech Inc v. Google LLC
I. Executive Summary and Procedural Information
- Parties & Counsel:
- Plaintiff: IPA Technologies Inc. (Delaware)
- Defendant: Google LLC (Delaware)
- Plaintiff’s Counsel: Skiermont Derby LLP; Bayard, PA.
- Case Identification: 1:18-cv-00318, D. Del., 02/26/2018
- Venue Allegations: Venue is alleged to be proper in the District of Delaware because Google LLC is incorporated in Delaware.
- Core Dispute: Plaintiff alleges that Defendant’s Google Assistant, Google Search, and related products infringe six patents related to software architectures for distributed electronic agents and speech-based data navigation.
- Technical Context: The technology concerns architectures for intelligent personal assistants, where a central "facilitator" coordinates tasks among various independent software "agents" to respond to complex user commands.
- Key Procedural History: The complaint alleges that Google was aware of the asserted patents and related technology years before the lawsuit was filed. Specifically, it alleges that several of the patents-in-suit were cited as prior art during the prosecution of Google's own patent applications. The complaint also notes that Google filed Inter Partes Review (IPR) petitions against several related patents in late 2017 and early 2018, shortly before this complaint was filed.
Case Timeline
| Date | Event |
|---|---|
| 1999-01-05 | Earliest Priority Date for all six Patents-in-Suit |
| 2003-02-18 | U.S. Patent No. 6,523,061 Issued |
| 2004-05-25 | U.S. Patent No. 6,742,021 Issued |
| 2004-06-29 | U.S. Patent No. 6,757,718 Issued |
| 2005-02-01 | U.S. Patent No. 6,851,115 Issued |
| 2006-04-25 | U.S. Patent No. 7,036,128 Issued |
| 2006-06-27 | U.S. Patent No. 7,069,560 Issued |
| 2014-10-21 | Alleged date of Google's knowledge of ’115 Patent via citation |
| 2016-05-01 | Google Assistant becomes publicly available |
| 2017-12-22 | Google files IPR petitions related to the ’061 Patent family |
| 2018-01-12 | Google files IPR petitions related to the ’021 and ’718 Patent families |
| 2018-02-26 | Complaint Filed |
II. Technology and Patent(s)-in-Suit Analysis
U.S. Patent No. 6,851,115 - "Software-based Architecture for Communication and Cooperation Among Distributed Electronic Agents"
The Invention Explained
- Problem Addressed: The patent’s background section describes that as computing systems became more complex and networked, traditional "direct manipulation" user interfaces (where a user masters a fixed set of commands for a single application) were becoming inadequate, particularly for non-expert users (’115 Patent, col. 2:21-33).
- The Patented Solution: The invention proposes a software architecture where autonomous "electronic agents" (specialized software programs) can cooperate to complete complex tasks for a user (’115 Patent, col. 1:25-29). This cooperation is managed by a central "facilitator," which is responsible for receiving a user's request, breaking it down into sub-tasks, and delegating those sub-tasks to the appropriate agents based on their registered capabilities (’115 Patent, col. 4:58-5:8). This architecture allows new agents to be added to the system dynamically, expanding its functionality without needing to rewrite the core code (’115 Patent, col. 8:41-55).
- Technical Importance: This agent-facilitator model provided a flexible framework for building complex, scalable, and user-friendly systems, such as early intelligent personal assistants, that could evolve over time by incorporating new services.
Key Claims at a Glance
- The complaint asserts independent claim 61 (Compl. ¶178).
- Claim 61 recites the essential elements of a "facilitator agent," which comprises:
- An "agent registry" that declares the capabilities of various service-providing electronic agents.
- A "facilitating engine" that can parse a service request to interpret a "compound goal" (a goal with multiple sub-parts).
- The service request is formed in an "Interagent Communication Language (ICL)" which includes both a "conversational protocol layer" (defining event types and parameters) and a "content layer" (defining goals, triggers, and data).
- The facilitating engine is further able to construct a "goal satisfaction plan" using various forms of reasoning and learning algorithms.
- The complaint reserves the right to assert additional claims (Compl. ¶179, n.10).
U.S. Patent No. 7,069,560 - "Highly Scalable Software-Based Architecture for Communication and Cooperation Among Distributed Electronic Agents"
The Invention Explained
- Problem Addressed: As a continuation of the ’115 Patent, the ’560 Patent addresses the same technical problem of managing complex tasks in a distributed computing environment (Compl. ¶52; ’115 Patent, col. 2:21-33).
- The Patented Solution: The ’560 Patent claims a computer-implemented process that uses the agent-facilitator architecture described in the ’115 Patent. The claimed process involves a service-requesting agent delegating sub-goals directly to other service-providing agents in a "peer to peer" manner, in addition to communicating through a central facilitator, to accomplish a larger "base goal" (’560 Patent, Claim 52).
- Technical Importance: This invention refines the agent-based architecture by detailing specific methods of task delegation, including peer-to-peer communication, which can improve efficiency in a distributed system.
Key Claims at a Glance
- The complaint asserts independent claim 52 (Compl. ¶207).
- Claim 52 recites a computer-implemented process comprising the steps of:
- Providing an agent registry.
- Interpreting a service request (a "base goal") formulated in an ICL.
- Determining the sub-goals needed to accomplish the base goal.
- Selecting a service-providing agent from the registry capable of completing a sub-goal.
- "delegating at least one sub goal as a peer to peer service request directly from a service requesting agent to a service providing agent".
- Delegating any remaining sub-goals to other selected agents.
- The complaint reserves the right to assert additional claims (Compl. ¶208, n.19).
U.S. Patent No. 7,036,128 - "Using a Community of Distributed Electronic Agents to Support a Highly Mobile Ambient Computing Environment"
- Technology Synopsis: This patent adapts the agent-facilitator architecture for a "highly mobile, ambient computing environment" (’128 Patent, col. 4:59-65). It specifically adds location-aware functionality, claiming a community of agents that includes "at least one location agent operable to ascertain a current physical location of a user" (’128 Patent, Claim 1).
- Asserted Claims: Independent claim 1 (Compl. ¶234).
- Accused Features: The complaint alleges that Google Assistant's ability to provide location-based services, such as directions, infringes this patent (Compl. ¶238-239).
U.S. Patent No. 6,742,021 - "Navigating Network-Based Electronic Information Using Spoken Input With Multimodal Error Feedback"
- Technology Synopsis: This patent focuses on improving speech-based navigation by using "multi-modal" feedback to resolve errors or ambiguities (’021 Patent, col. 1:22-26). Instead of simply failing when a spoken request is unclear, the system solicits additional input from the user via a "non-spoken modality" (e.g., a touch-screen menu) to refine the query (’021 Patent, Claim 1).
- Asserted Claims: Independent claim 1 (Compl. ¶259).
- Accused Features: The complaint accuses Google Assistant's process of receiving a spoken query and then presenting visual, tappable options on a screen as infringing this patent (Compl. ¶262-265).
U.S. Patent No. 6,523,061 - "System, Method, and Article of Manufacture For Agent-Based Navigation in a Speech-Based Data Navigation System"
- Technology Synopsis: This patent claims a method for speech-based navigation that explicitly utilizes the agent-based architecture. A spoken request is interpreted and used to construct a query, which is then routed to "at least one agent" for execution, with a "facilitator" managing the data flow and maintaining a registration of each agent's capabilities (’061 Patent, Claim 1).
- Asserted Claims: Independent claim 1 (Compl. ¶282).
- Accused Features: The complaint alleges that Google Assistant's use of a facilitator (the Assistant itself) to manage various agents (e.g., Search, phone dialer) in response to a spoken request infringes this patent (Compl. ¶285-291).
U.S. Patent No. 6,757,718 - "Mobile Navigation of Network-Based Electronic Information Using Spoken Input"
- Technology Synopsis: This patent claims a method for speech-based navigation specifically using a "mobile information appliance" defined as a "portable remote control device or a set-top box for a television" (’718 Patent, Claim 1). The method involves receiving a spoken request on this appliance and transmitting data from a network server to it.
- Asserted Claims: Independent claim 1 (Compl. ¶305).
- Accused Features: The complaint alleges that Google Assistant's functionality on devices like Google Home and Chromecast, which control televisions via spoken commands, infringes this patent (Compl. ¶308-310).
III. The Accused Instrumentality
Product Identification
- Google Assistant, Google Search, Google Actions, Dialogflow, and related products and services (Compl. ¶164, ¶169, ¶178).
Functionality and Market Context
- Google Assistant is a voice-activated intelligent personal assistant that operates on a wide range of devices, including smartphones, smart speakers, and televisions (Compl. ¶165-166). It functions by interpreting a user's natural language request and coordinating with various "electronic agents" to fulfill the request (Compl. ¶167).
- These agents include both Google's own services (like Search) and a large ecosystem of third-party applications and services available through the "Actions on Google" platform and the "Assistant app directory" (Compl. ¶168, ¶173). The complaint alleges that Google Assistant acts as a central coordinator, or facilitator, that determines a user's intent and invokes the appropriate agent to complete the task (Compl. ¶181). This process is illustrated in a diagram from Google's developer documentation, which shows the Assistant querying "Actions on Google" to find a suitable app for a user's request (Compl. p. 61).
IV. Analysis of Infringement Allegations
U.S. Patent No. 6,851,115 Infringement Allegations
| Claim Element (from Independent Claim 61) | Alleged Infringing Functionality | Complaint Citation | Patent Citation |
|---|---|---|---|
| A facilitator agent arranged to coordinate cooperative task completion within a distributed computing environment... | Google Assistant allegedly acts as the facilitator agent, coordinating tasks among third-party apps and other electronic agents (Compl. ¶181). | ¶181 | col. 4:62-65 |
| an agent registry that declares capabilities of service-providing electronic agents... | The "Actions on Google" platform and the "Assistant app directory" allegedly function as a registry where third-party developers declare the capabilities of their agent apps (Compl. ¶182). | ¶182 | col. 7:2-4 |
| a facilitating engine operable to parse a service requesting order to interpret a compound goal... | Google Assistant's natural language processing allegedly parses a user's request to determine their intent (the goal) and invokes the best app to fulfill it (Compl. ¶183-184). | ¶183 | col. 4:65-5:8 |
| ...the service request formed according to an Interagent Communication Language (ICL), wherein the ICL includes: a layer of conversational protocol defined by event types and parameter lists...and a content layer comprising one or more of goals... | Google's DialogFlow suite, which defines "Intents" (goals), "event types," and associated "parameters," is alleged to be the ICL used for communication between the Assistant and its agents (Compl. ¶186-188). A screenshot from DialogFlow shows the definition of "Action & parameters" (Compl. p. 68). | ¶186 | col. 10:48-55 |
| the facilitating engine further operable to construct a goal satisfaction plan by using reasoning that includes...domain-independent coordination strategies...and application-specific reasoning comprising rules and learning algorithms. | Google Assistant's "recommendation algorithm" allegedly determines which app can complete a task, and its use of machine learning and "Training Phrases" to define intents constitutes the claimed reasoning and learning algorithms (Compl. ¶190-191). | ¶190 | col. 11:37-43 |
Identified Points of Contention
- Scope Questions: A central question may be whether the Google Assistant platform, where a primary software entity (Assistant) communicates with third-party applications via APIs, constitutes a "distributed computing environment having a plurality of autonomous service-providing electronic agents" as described in the patent. The defense may argue that "Actions on Google" are services called by a primary application rather than "autonomous agents" cooperating in a distributed system.
- Technical Questions: The analysis may turn on whether Google's system of "Intents" and "parameters" in its DialogFlow API meets the specific two-layer definition of an "Interagent Communication Language (ICL)" having distinct "conversational protocol" and "content" layers as claimed.
U.S. Patent No. 7,069,560 Infringement Allegations
| Claim Element (from Independent Claim 52) | Alleged Infringing Functionality | Complaint Citation | Patent Citation |
|---|---|---|---|
| providing at least one agent registry including capabilities of service providing electronic agents; | As with the ’115 Patent, the "Actions on Google" platform is alleged to be the agent registry (Compl. ¶212). | ¶212 | col. 7:2-4 |
| interpreting a service request in the form of a base goal, the service request being in a interagent communication language (ICL)... | Google Assistant allegedly interprets the user's spoken request (the base goal) using its natural language processing capabilities, which operate via the alleged ICL (Compl. ¶214). A screenshot shows Google interpreting a request for transit directions (Compl. p. 79). | ¶214 | col. 10:48-55 |
| determining a plurality of sub goals necessary to accomplish the base goal; | The complaint alleges that as part of defining "intents" for Actions on Google, "Training Phrases" are used in machine learning to determine the sub-goals needed to identify a user's request (Compl. ¶217). | ¶217 | col. 11:37-43 |
| selecting from said registry at least one service providing agent capable of completing said sub goals; | A "recommendation algorithm" within Actions on Google allegedly determines and selects an app (a service providing agent) that has an action capable of completing the user's task (Compl. ¶218). | ¶218 | col. 11:13-16 |
| delegating at least one sub goal as a peer to peer service request directly from a service requesting agent to a service providing agent; | The complaint alleges that Actions on Google delegates the sub-goal directly from the service requesting agent (Google Assistant) to the service providing agent (the selected app) for fulfillment (Compl. ¶219). | ¶219 | col. 14:14-24 |
Identified Points of Contention
- Scope Questions: A key dispute may arise over the term "peer to peer service request." The infringement theory suggests a request from Google Assistant to a third-party app is "peer to peer." The defense may argue this is a standard, hierarchical client-server API call, and that "peer to peer" in the context of the patent requires communication between two non-facilitator agents without the facilitator's direct involvement.
- Technical Questions: It raises a question of whether the process of an integrated system (Google Assistant) calling a function in an external application (an "Action") is equivalent to the claimed process of a "service requesting agent" delegating a request to a distinct "service providing agent."
V. Key Claim Terms for Construction
"facilitator agent" (’115 Patent, claim 61)
- Context and Importance: This term is the central component of the claimed architecture. The outcome of the case may depend on whether Google Assistant itself is properly characterized as a "facilitator agent" that coordinates other "autonomous" agents, or if it is simply a monolithic application that calls external services.
- Intrinsic Evidence for Interpretation:
- Evidence for a Broader Interpretation: The specification states a facilitator is "responsible for matching requests, from users and agents, with descriptions of the capabilities of other agents" (’115 Patent, col. 4:62-65). This functional description could support an argument that any entity performing this matching and delegating role, like Google Assistant, is a facilitator agent.
- Evidence for a Narrower Interpretation: Figure 4 of the patent depicts the "Facilitator Agent" as a distinct architectural block, separate from the "User Interface Agent" (’115 Patent, Fig. 4). This could support an argument that the facilitator must be architecturally separate from the user-facing component, a distinction that may not exist in the integrated Google Assistant product.
"Interagent Communication Language (ICL)" (’115 Patent, claim 61)
- Context and Importance: The infringement case hinges on mapping Google's developer tools (DialogFlow, Actions SDK) to this claimed term. Practitioners may focus on whether a modern, web-based API framework falls within the scope of the specific two-layer "language" defined in the patent.
- Intrinsic Evidence for Interpretation:
- Evidence for a Broader Interpretation: Claim 61 defines the ICL functionally as having a "conversational protocol layer" (events, parameters) and a "content layer" (goals, triggers). Plaintiff may argue that Google's system of defining "Intents" (goals) and "parameters" fits this functional definition (Compl. ¶187-188).
- Evidence for a Narrower Interpretation: The specification provides detailed examples of the ICL based on the PROLOG programming language (’115 Patent, col. 10:48-61). The defense may argue that the term "ICL" is limited to a language with this specific logic-based structure, which differs from Google's JSON/HTTP-based APIs.
VI. Other Allegations
- Indirect Infringement: The complaint alleges active inducement, stating that Google encourages and provides instructions to end-users and developers on how to use the accused systems in an infringing manner (Compl. ¶197-198, ¶225).
- Willful Infringement: The complaint alleges willful infringement based on pre-suit knowledge. It specifically alleges that Google gained knowledge of the ’115 Patent no later than October 21, 2014, when the patent was cited by a USPTO Examiner during the prosecution of Google's own U.S. Patent No. 9,031,830 (Compl. ¶195, ¶74). Similar allegations are made for the other asserted patents based on citations in Google's patent prosecution files (Compl. ¶248, ¶269, ¶293).
VII. Analyst’s Conclusion: Key Questions for the Case
- A core issue will be one of architectural equivalence: does Google's centrally-managed ecosystem, where the Google Assistant application makes API calls to third-party "Actions," embody the "distributed computing environment" of "autonomous" agents coordinated by a distinct "facilitator agent" as claimed in the patents, or is this a fundamentally different client-server model?
- A second key issue will be one of definitional scope: can the term "Interagent Communication Language (ICL)," which the patent describes with reference to logic-based programming constructs, be construed to cover the modern web APIs, SDKs, and data formats (like JSON) that govern the interaction between Google Assistant and third-party developers?
- A third critical question will concern willfulness: what is the legal effect of the specific allegations that Google was made aware of the patents-in-suit during the prosecution of its own patent portfolio, years before the litigation began? This could be a central factor in determining the extent of any potential damages.