DCT

2:25-cv-00918

Honeywell Intl Inc v. Eps Ehrhardt + Partner Solutions Inc

Log In

I. Executive Summary and Procedural Information

  • Parties & Counsel:
  • Case Identification: 2:25-cv-00918, E.D. Tex., 08/29/2025
  • Venue Allegations: Plaintiff alleges venue is proper in the Eastern District of Texas because Defendant maintains its principal place of business in Northlake, Texas, which is within the district.
  • Core Dispute: Plaintiff alleges that Defendant’s warehouse logistics and voice-picking products infringe seven patents related to voice-enabled workflow management, 3D model annotation, and logistics optimization.
  • Technical Context: The dispute centers on technology for warehouse automation and logistics management, a field critical to improving the speed and accuracy of modern supply chains and order fulfillment.
  • Key Procedural History: The complaint details extensive pre-suit correspondence between the parties spanning from January 2022 to August 2024, during which Plaintiff repeatedly notified Defendant of alleged infringement across a growing number of patents. The complaint also notes that on July 9, 2025, Defendant filed a declaratory judgment action against Plaintiff in the Western District of North Carolina regarding the patents at issue.

Case Timeline

Date Event
2008-11-14 ’261 Patent Priority Date
2011-05-24 ’940 Patent Priority Date
2013-02-26 ’261 Patent Issue Date
2014-08-26 ’504 and ’863 Patents Priority Date
2015-10-06 ’940 Patent Issue Date
2016-07-27 ’336, ’139, and ’678 Patents Priority Date
2018-08-14 ’504 Patent Issue Date
2020-09-08 ’863 Patent Issue Date
2021-10-26 ’336 Patent Issue Date
2022-01-18 Plaintiff sends first notice letter to Defendant
2023-05-02 Plaintiff sends second notice letter to Defendant
2024-06-23 Plaintiff sends notice letter including ’261 Patent
2024-08-14 Parties’ business representatives meet
2024-08-06 ’139 Patent Issue Date
2025-07-09 Defendant files declaratory judgment action in W.D.N.C.
2025-08-26 ’678 Patent Issue Date
2025-08-29 Complaint Filing Date

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

U.S. Patent No. 11,158,336 - "Distinguishing user speech from background speech in speech-dense environments"

  • Issued: October 26, 2021

The Invention Explained

  • Problem Addressed: The patent describes the challenge of operating voice-controlled systems in noisy industrial environments, such as warehouses or logistics centers, where extraneous sounds interfere with recognizing a user's spoken commands (US 11,158,336 B2, col. 2:30-44). Specifically, it highlights the difficulty of distinguishing the user's voice from background conversations and announcements from public address (PA) systems, which share the aural qualities of intended speech input (US 11,158,336 B2, col. 2:45-54).
  • The Patented Solution: The invention proposes a method and system that prepares a voice-data model in advance of field use by creating it in a training environment that includes both the desired user's speech and unwanted background sounds (US 11,158,336 B2, Abstract). This "training corpus" is used to train a speech recognition system to identify the intended user's speech, even when spoken concurrently with background noise, by using a processor-based learning system to duplicate assessments made by human listeners who "tag" the pre-recorded sounds (US 11,158,336 B2, col. 3:5-17; Fig. 4).
  • Technical Importance: This approach allows for the creation of more robust voice recognition models tailored to specific, high-noise environments, potentially improving accuracy and reliability where standard noise cancellation is insufficient (US 11,158,336 B2, col. 11:46-53).

Key Claims at a Glance

  • The complaint asserts independent claim 17 (Compl. ¶63).
  • Essential elements of claim 17 (a non-transitory computer-readable medium) include instructions for:
    • Receiving an audio input.
    • Determining if the input matches an "expected verbalization."
    • If it matches, categorizing the input as "user speech."
    • If it does not match, comparing the input with an "audio template" that comprises characterizations of user speech samples, background speech samples, and background environmental sound samples.
    • Based on the comparison, determining the input to be either "user speech" or "background speech."
  • The complaint reserves the right to assert additional claims (Compl. ¶63).

U.S. Patent No. 8,386,261 - "Training/coaching system for a voice-enabled work environment"

  • Issued: February 26, 2013

The Invention Explained

  • Problem Addressed: The patent addresses the difficulty of training users on voice-enabled workflow systems, because a trainer cannot hear the specific voice prompts the system is delivering to the trainee (US 8,386,261 B2, col. 2:20-29). Prior hardware solutions, such as connecting a separate loudspeaker to the trainee's device, are described as "cumbersome to use" and requiring "direct (and obtrusive) interaction" (US 8,386,261 B2, col. 2:32-38).
  • The Patented Solution: The invention provides a system where the voice prompts delivered to a first user's (trainee's) voice assistant are also wirelessly communicated to the voice assistant of a second user (coach) (US 8,386,261 B2, Abstract). The coach's device enters a "Coaching mode" where it receives system prompts from the trainee's device over the wireless network, allowing the coach to hear exactly what the trainee hears in real-time without physical tethers or extra hardware (US 8,386,261 B2, col. 2:50-57; Fig. 3).
  • Technical Importance: This technology aims to make training on voice-directed systems more effective and less intrusive by allowing for remote, real-time monitoring of the user-system interaction (US 8,386,261 B2, col. 3:3-7).

Key Claims at a Glance

  • The complaint asserts independent claim 6 (Compl. ¶76).
  • Essential elements of claim 6 (a device) include:
    • A network interface for wireless communication.
    • A voice user interface for converting system prompts to speech and user voice commands to system commands.
    • A "coaching mode module" that can be activated to coach another device.
    • When activated, the module is operable to establish a connection with the other device over the wireless network.
    • The module is further operable to receive system prompts from the other device over that connection and provide them to its own voice user interface for output.
  • The complaint reserves the right to assert additional claims (Compl. ¶76).

Multi-Patent Capsule: U.S. Patent No. 9,152,940

  • Patent Identification: U.S. Patent No. 9,152,940, "Method and apparatus for optimized shipping strategies accounting for endpoint requirements," issued October 6, 2015.
  • Technology Synopsis: The patent addresses inefficiencies in warehouse order fulfillment that arise because products are picked based on their location in the distribution center, not their final placement location in a retail store (US 9,152,940 B2, col. 1:11-26). The invention provides a method for "cartonization"—organizing items into shipping containers—that accounts for the end-store's layout (planogram), thereby optimizing the subsequent shelf-stocking process (US 9,152,940 B2, col. 2:43-51).
  • Asserted Claims: The complaint asserts at least claim 7 (Compl. ¶89).
  • Accused Features: The accused features are those within Defendant's warehouse management systems (WMS / LFS and EPS One Suite) that allegedly organize and direct order picking and packing (Compl. ¶87).

Multi-Patent Capsule: U.S. Patent No. 10,049,504

  • Patent Identification: U.S. Patent No. 10,049,504, "Annotating three-dimensional displays," issued August 14, 2018.
  • Technology Synopsis: The patent addresses the problem of text annotations on 3D models (e.g., building information models, or BIMs) becoming cluttered, unreadable, or occluded when the user rotates or zooms the view (US 10,049,504 B2, col. 1:38-48). The solution is a system that intelligently manages the display of annotations based on context, such as user interaction or system alerts, and can automatically offset or dim labels to avoid overlap and maintain visual clarity (US 10,049,504 B2, col. 2:12-21, 59-64).
  • Asserted Claims: The complaint asserts at least claim 1 (Compl. ¶103).
  • Accused Features: The accused features are part of Defendant's Warehouse Control System (WCS) and EPS ONE products, which allegedly use annotated 3D displays for managing warehouse operations (Compl. ¶101).

Multi-Patent Capsule: U.S. Patent No. 10,769,863

  • Patent Identification: U.S. Patent No. 10,769,863, "Annotating three-dimensional displays of a particular view of a 3D model," issued September 8, 2020.
  • Technology Synopsis: Related to the ’504 Patent, this invention also focuses on improving the readability of annotations on 3D models of facilities (US 10,769,863 B2, col. 1:36-49). It describes methods for displaying a subset of annotations based on a determined context and for modifying their display—such as by offsetting them and linking them with a line to their corresponding object—to prevent visual clutter and occlusion as the user manipulates the 3D view (US 10,769,863 B2, col. 2:53-66).
  • Asserted Claims: The complaint asserts at least claim 1 (Compl. ¶117).
  • Accused Features: The accused features are within Defendant's Warehouse Management Software, which allegedly uses these 3D annotation techniques (Compl. ¶115).

Multi-Patent Capsule: U.S. Patent No. 12,057,139

  • Patent Identification: U.S. Patent No. 12,057,139, "Distinguishing user speech from background speech in speech-dense environments," issued August 6, 2024.
  • Technology Synopsis: This patent is from the same family as the ’336 Patent and addresses the same technical challenge: enabling accurate voice recognition in noisy industrial settings (US 12,057,139 B2, col. 2:30-44). The invention involves pre-field-use creation of a voice model using a training corpus that includes both user speech and various background sounds, which is then used by a learning system to distinguish the intended user's voice from ambient noise during operation (US 12,057,139 B2, Abstract).
  • Asserted Claims: The complaint asserts at least claim 28 (Compl. ¶130).
  • Accused Features: The accused products are Defendant’s "VoiceWear Suite," which includes the Lydia VoiceWear hardware, Voxter computer, and Lydia Voice software (Compl. ¶129).

Multi-Patent Capsule: U.S. Patent No. 12,400,678

  • Patent Identification: U.S. Patent No. 12,400,678, "Distinguishing user speech from background speech in speech-dense environments," issued August 26, 2025.
  • Technology Synopsis: Also from the same family as the ’336 and ’139 patents, this patent describes a system for improving speech recognition accuracy where extraneous background noise, including other conversations and PA system announcements, is prevalent (US 12,400,678 B2, col. 2:30-54). The solution involves training a recognition model on a corpus of audio that includes both desired user speech and unwanted background noise, enabling the system to better isolate and interpret user commands in the field (US 12,400,678 B2, col. 3:5-17).
  • Asserted Claims: The complaint asserts at least claim 1 (Compl. ¶142).
  • Accused Features: The complaint accuses Defendant’s Lydia voice software of infringing this patent (Compl. ¶141).

III. The Accused Instrumentality

Product Identification

  • The complaint accuses a suite of hardware and software products from EPS, including: EPS's Lydia Voice software, Warehouse Management System (WMS/LFS), Warehouse Control System (WCS), EPG ONE Supply Chain Execution Suite, Voxter mobile computers (e.g., Voxter Vantage), and Lydia VoiceWear wearable vests (Compl. ¶23).

Functionality and Market Context

  • The accused instrumentalities form a comprehensive system for managing warehouse logistics and enabling voice-directed workflows (Compl. ¶24, ¶25).
  • The "Lydia Voice" software is described as a "leading voice picking solution" that guides workers through tasks (Compl. ¶24).
  • Hardware such as the Voxter Vantage is a "mobile voice computer" designed to be used with the Lydia Voice software (Compl. ¶26).
  • The Lydia VoiceWear products are "unique picking vest[s]" with integrated speakers and microphones that provide a hands-free interface for workers (Compl. ¶27). The complaint provides a marketing image showing the wearable "LYDIA VoiceWear Air" and "LYDIA VoiceWear 5" vests (Compl. p. 7).
  • The WMS, WCS, and EPG ONE software suites are alleged to provide backend control, guiding the movement of materials and empowering warehouse automation (Compl. ¶24, ¶25).

IV. Analysis of Infringement Allegations

U.S. Patent No. 11,158,336 Infringement Allegations

Claim Element (from Independent Claim 17) Alleged Infringing Functionality Complaint Citation Patent Citation
[a non-transitory computer readable medium comprising a set of instructions that are executable by a processor to:] The Lydia Voice software is a non-transitory medium containing executable instructions. ¶62 col. 16:17-19
receive an audio input; The Lydia Voice system receives audio input from a user via a microphone on a headset or wearable vest. ¶27 col. 4:11-14
determine if the received audio input is a vocalization that matches an expected verbalization, The Lydia Voice software is alleged to process received audio to determine if it corresponds to expected commands or "hints" in a given workflow context. ¶62 col. 19:26-30
in response to determining that the received audio input matches an expected verbalization, categorize the received audio input as a user speech When a match to an expected command is found, the system allegedly categorizes it as valid user speech for processing. ¶62 col. 19:31-34
in response to determining that the received audio input does not match any of the expected verbalizations, compare the audio input with an audio template comprising audio characterizations of a plurality of sound samples comprising at least one of user speech samples, background speech samples, background environmental sound samples; The complaint alleges that when the audio is not an expected command, Lydia Voice compares it against a pre-trained model ("audio template") that was developed using samples of both user speech and background noise to distinguish between them. ¶62 col. 11:25-45
and determine that the received audio input to be one of: the user speech and a background speech. Based on this comparison, the Lydia Voice system determines whether the audio is legitimate user speech to be acted upon or background noise to be ignored. ¶62 col. 11:17-24

Identified Points of Contention

  • Technical Questions: A central technical question may be how Defendant's Lydia Voice system actually performs noise filtering. Does it use an "audio template" that contains stored "characterizations" of distinct user speech, background speech, and environmental sound samples as required by the claim, or does it employ a different type of machine learning model or signal processing technique that does not map to these claim elements?
  • Scope Questions: The dispute may raise the question of whether the functionality for recognizing "expected verbalizations" (or "hints") is distinct from the general comparison against the "audio template." The analysis will likely focus on whether the accused system performs the sequential, conditional logic required by the claim language.

U.S. Patent No. 8,386,261 Infringement Allegations

Claim Element (from Independent Claim 6) Alleged Infringing Functionality Complaint Citation Patent Citation
[A device, comprising:] a network interface operable to communicate with a wireless network; The accused Lydia Voice system runs on mobile computers that include wireless networking interfaces (e.g., Wi-Fi, Bluetooth). ¶75 col. 4:39-44
a voice user interface operable to, upon receiving a system Prompt, convert the system Prompt into speech...and, upon receiving a voice command from the user, use speech recognition technology... The Lydia Voice system provides voice prompts to users and accepts voice commands from them to direct workflows. ¶24 col. 2:20-29
the device having a coaching mode module...operable to be activated into a coaching mode to coach the user of at least one other device, The complaint alleges that the Lydia Voice system on mobile computers includes functionality that allows one user (a coach) to monitor the prompts being sent to another user (a trainee). ¶75 col. 2:50-52
the coaching mode module...further operable to obtain a network address for the at least one other device and to establish a connection with the [other device] over a wireless network... The accused system allegedly establishes a wireless network connection between the coach's device and the trainee's device to enable the monitoring functionality. ¶75 col. 3:3-7
the coaching mode module operable to receive system Prompts from the at least one other device...and to provide the system Prompts for the voice user interface to be output as a voice Prompt for the user of the device. The accused system allegedly transmits the system prompts intended for the trainee's device over the network to the coach's device, which then outputs them as audio for the coach to hear. ¶75 col. 9:4-10

Identified Points of Contention

  • Scope Questions: A primary point of contention will likely be the definition of a "coaching mode module." Does the accused system contain a specific, dedicated software module for training that performs the claimed functions, or does it use a more generic audio-sharing or screen-mirroring capability that Plaintiff will argue is equivalent? The question for the court will be whether the accused functionality meets the specific structural and functional requirements of this claim term.
  • Technical Questions: The analysis may turn on the specific mechanism of communication. Does the accused system "receive system Prompts" from the trainee's device (which could imply receiving the underlying data or text of the prompt) and then locally synthesize the audio on the coach's device, as the claim suggests? Or does it simply stream the audio output from the trainee's device to the coach's device, which could represent a different technical implementation?

V. Key Claim Terms for Construction

For the ’336 Patent

  • The Term: "audio template" (from claim 17)
  • Context and Importance: This term appears to be the core of the invention's method for distinguishing user speech from noise. Its construction will be critical because the infringement analysis depends on whether Defendant's machine learning model in Lydia Voice can be characterized as an "audio template" that "compris[es] audio characterizations of...user speech samples, background speech samples, [and] background environmental sound samples." Practitioners may focus on whether this requires a stored library of discrete, pre-classified sound profiles or if it can read on the weighted parameters of an integrated neural network.
  • Intrinsic Evidence for Interpretation:
    • Evidence for a Broader Interpretation: The specification refers to the template as an "audio characterization model ACM" derived from a "training corpus TC" (US 11,158,336 B2, col. 10:1-5). This language may support a broader interpretation where the "template" is the resulting trained model itself, not necessarily a simple collection of raw samples.
    • Evidence for a Narrower Interpretation: The claim itself breaks the template down into constituent parts ("user speech samples, background speech samples, background environmental sound samples"). This explicit listing could support a narrower construction requiring the model to be built from or contain distinguishable characterizations of these specific sound categories, rather than a more holistic, abstract model.

For the ’261 Patent

  • The Term: "coaching mode module" (from claim 6)
  • Context and Importance: This term defines the central component of the claimed training system. The dispute will likely hinge on whether the accused Lydia Voice system has a software component that meets the definition of this "module." The construction will determine if any feature allowing a supervisor to listen in on a user's session infringes, or if only a feature with the specific architecture described in the patent (e.g., establishing a dedicated connection to forward prompts) meets the claim limitation.
  • Intrinsic Evidence for Interpretation:
    • Evidence for a Broader Interpretation: The Summary of the Invention describes the invention more generally as directing "voice prompts delivered to a first user...to also be communicated wirelessly to the voice assistant device of a second user" (US 8,386,261 B2, col. 2:50-54). This could suggest that the "module" is any software that achieves this outcome.
    • Evidence for a Narrower Interpretation: The detailed description explains that the "Coaching mode module 120 sends a look-up table request...to obtain the local network IP address" and "opens or establishes a direct socket connection 150" (US 8,386,261 B2, col. 7:50-59). This detailed description of establishing a specific type of connection to forward prompts (not just raw audio) may support a narrower construction requiring these specific functionalities.

VI. Other Allegations

  • Indirect Infringement: The complaint alleges that Defendant induces infringement by providing customers with marketing materials, instructional materials, use case studies, and YouTube videos that instruct and encourage end-users to operate the accused products in a manner that infringes the asserted patents (Compl. ¶67, ¶80, ¶93). The complaint also alleges contributory infringement, stating the products are especially made for infringing use (Compl. ¶68, ¶81, ¶94).
  • Willful Infringement: The willfulness allegations are based on Defendant’s alleged knowledge of the patents and infringement dating back to at least January 18, 2022, when Plaintiff sent its first notice letter (Compl. ¶32, ¶44). The complaint details a multi-year history of correspondence in which Plaintiff identified the patents and its infringement theories, alleging that Defendant's continued activities despite this knowledge constitute willful infringement (Compl. ¶69, ¶82, ¶95).

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

  • A core issue will be one of technical implementation: For the patents on distinguishing user speech ('336, '139, '678), does Defendant's Lydia Voice system utilize a trained "audio template" built from discrete user and background sound characterizations as claimed, or does it employ a fundamentally different machine learning or signal processing architecture for noise cancellation that falls outside the patent's scope?
  • A central question of functional scope will arise for the training system patent ('261): does the supervisory monitoring feature in Defendant's products constitute the specific "coaching mode module" recited in the claims—which requires establishing a network connection to forward system prompts—or is it a more generic audio-streaming function that operates in a technically distinct manner?
  • A key evidentiary question will concern willfulness: Given the extensive pre-suit notice alleged in the complaint, the case will likely turn on the objective reasonableness of Defendant's non-infringement and invalidity defenses developed both before and after the suit was filed. The resolution of this issue will be highly dependent on the evidence presented regarding Defendant's state of mind and the advice it received.