DCT
1:25-cv-00476
All Terminal Services LLC v. Roboflow Inc
I. Executive Summary and Procedural Information
- Parties & Counsel:
- Plaintiff: ALL TERMINAL SERVICES, LLC D/B/A CONGLOBAL TECHNOLOGIES (Illinois)
- Defendant: Roboflow, Inc. (Delaware)
- Plaintiff’s Counsel: Potter Anderson & Corroon LLP; Arnall Golden Gregory LLP
- Case Identification: 1:25-cv-00476, D. Del., 04/18/2025
- Venue Allegations: Plaintiff alleges venue is proper in the District of Delaware because Defendant is a corporation organized and existing under the laws of the State of Delaware.
- Core Dispute: Plaintiff alleges that Defendant’s AI-powered Yard Management System infringes three patents related to the automated visual identification, classification, and tracking of assets in rail yards and intermodal terminals.
- Technical Context: The technology addresses automated inventory management in logistics, a sector where real-time asset visibility is critical for supply chain efficiency.
- Key Procedural History: The complaint alleges that Plaintiff provided Defendant with written notice of infringement of the ’148 Patent via a letter dated December 3, 2024, which may be relevant to the allegations of willful infringement.
Case Timeline
| Date | Event |
|---|---|
| 2019-11-18 | Priority Date for ’148, ’183, and ’439 Patents |
| 2024-06-25 | ’148 Patent Issued |
| 2024-12-03 | Plaintiff sent notice letter to Defendant regarding ’148 Patent |
| 2024-12-04 | Defendant received notice letter regarding ’148 Patent |
| 2025-02-04 | ’183 Patent Issued |
| 2025-03-18 | ’439 Patent Issued |
| 2025-04-18 | Complaint Filed |
II. Technology and Patent(s)-in-Suit Analysis
U.S. Patent No. 12,020,148 - Control System for Railway Yard and Related Methods
The Invention Explained
- Problem Addressed: The patent’s background section notes that tracking shipping containers and railcars within a railway yard is difficult due to the large number of tracks and trains (’148 Patent, col. 1:33-38). The complaint further elaborates that automated tracking using traditional OCR is unreliable because shipping assets are often dirty, damaged, or miscolored, leading to inaccurate imaging (Compl. ¶22).
- The Patented Solution: The invention is a control system for a railway yard that uses "railyard sensors" (such as cameras) to collect visual and other data about railcars on the tracks. A server processes this data, using technologies like OCR and machine learning, to generate a database that identifies and classifies each railcar. This information is then used to "selectively control" remote control locomotives (RCLs) to position the railcars within the yard, automating inventory management and movement (’148 Patent, Abstract; col. 2:1-15).
- Technical Importance: The technology provides a technological advancement over prior art methods by using a novel combination of cameras, algorithms, and machine learning to more accurately and efficiently identify and manage yard assets (Compl. ¶21).
Key Claims at a Glance
- The complaint asserts independent Claim 1 (Compl. ¶48).
- Essential elements of Claim 1 include:
- A control system for a railway yard comprising a plurality of remote control locomotives (RCLs) and sets of railcars.
- A plurality of railyard sensors configured to generate railyard sensor data.
- A server in communication with the RCLs and sensors.
- The server is configured to generate a database associated with the railcars based on the sensor data.
- The server is configured to selectively control the RCLs to position the sets of railcars based on the sensor data.
- The complaint does not explicitly reserve the right to assert dependent claims.
U.S. Patent No. 12,217,183 - Control System for Railway Yard and Related Methods
The Invention Explained
- Problem Addressed: The patent addresses the same technical challenges as the ’148 Patent: the difficulty of tracking assets in a complex railway yard environment (’183 Patent, col. 1:21-39).
- The Patented Solution: This invention focuses on the server component of the control system. It claims a server comprising processors and memory storing algorithms that, when executed, perform the key control functions: generating a database from sensor data, identifying individual railcars using image data, and selectively controlling RCLs to position the railcars based on that data (’183 Patent, Abstract; col. 2:16-30). This shifts the inventive focus from the overall system to the specific programming and operation of the central server.
- Technical Importance: This approach centralizes the system's intelligence, enabling sophisticated, algorithm-driven management of yard operations from a single control point (Compl. ¶21).
Key Claims at a Glance
- The complaint asserts independent Claim 16 (Compl. ¶56).
- Essential elements of Claim 16 include:
- A server for controlling a railway yard comprising one or more processors and non-transitory computer-readable storage mediums.
- The mediums store algorithms that cause the processors to perform steps.
- The steps include generating a database associated with railcars based on railyard sensor data.
- The steps include selectively controlling one or more RCLs to position the railcars based on the sensor data.
- The steps include identifying each railcar based upon railyard image data.
- The complaint does not explicitly reserve the right to assert dependent claims.
U.S. Patent No. 12,254,439 - Control System for Container Terminal and Related Methods
- Patent Identification: U.S. Patent No. 12,254,439, "Control System for Container Terminal and Related Methods," issued March 18, 2025 (Compl. ¶15).
- Technology Synopsis: This patent broadens the scope of the inventions from a "railway yard" to a more general "inventory management facility" or "container terminal." It claims a control system that uses sensors to generate data about "shipping assets," a server to create a tracking database, machine learning, and a vehicle equipped with a geolocation device, a wireless transceiver, and a controller (Compl. ¶64).
- Asserted Claims: Independent Claim 1 (Compl. ¶64).
- Accused Features: The complaint accuses Roboflow's Yard Management System, alleging it uses vehicle-mounted cameras, OCR, machine learning, and database creation to track shipping assets in a manner that infringes the claims (Compl. ¶¶65-66).
III. The Accused Instrumentality
Product Identification
- The "Accused Systems" include Roboflow's Yard Management System, Workflow system, and Roboflow Interference (Compl. ¶4).
Functionality and Market Context
- The Accused Systems are alleged to be an AI-driven platform for managing inventory in intermodal and rail yards (Compl. ¶26). The system reportedly uses cameras, which may be mounted on vehicles like hostler trucks, to capture continuous image streams of yard assets such as containers and chassis (Compl. ¶¶27, 33). These images are then processed using computer vision, optical character recognition (OCR), and machine learning models to automatically extract identifiers (e.g., container numbers, chassis IDs), logos, and other asset information (Compl. ¶¶28, 33). This extracted data is organized into a digital database to provide users with real-time visibility into asset location and status, enabling automated tracking and classification (Compl. ¶28). The complaint includes a screenshot from Roboflow's website illustrating the system identifying and labeling attributes of intermodal containers in a yard (Compl. ¶34, p. 7).
IV. Analysis of Infringement Allegations
’148 Patent Infringement Allegations
| Claim Element (from Independent Claim 1) | Alleged Infringing Functionality | Complaint Citation | Patent Citation |
|---|---|---|---|
| a plurality of railyard sensors configured to generate railyard sensor data of the plurality of railroad tracks | The Accused Systems use vehicle-mounted cameras to collect image data from yard assets, which function as railyard sensors. | ¶50 | col. 3:41-43 |
| a server in communication with... said plurality of railyard sensors | Roboflow's systems operate a server that receives and processes the image data collected by the cameras. | ¶¶48, 58 | col. 3:48-54 |
| said server configured to generate a database associated with the sets of railcars based upon the railyard sensor data | The Accused Systems apply OCR and machine learning to the image data to extract information like container numbers and logos, and organize this information into a structured digital database. | ¶50 | col. 2:23-28 |
| said server configured to... selectively control said plurality of RCLs to position the sets of railcars within the plurality of railroad tracks based upon the railyard sensor data | The complaint makes a conclusory allegation that the Accused Systems infringe this element, but does not provide specific factual support for how Roboflow's system controls RCLs. | ¶48 | col. 2:26-30 |
Identified Points of Contention
- Scope Questions: A central question may be whether Defendant "makes, uses, [or] sells" the entire claimed "control system," which includes physical hardware like "railyard sensors" and "RCLs." The complaint alleges the Accused Systems are used "in combination with" such hardware (Compl. ¶30), raising potential questions of divided infringement and whether Defendant directs or controls its customers' actions to the extent required to establish liability for the entire system.
- Technical Questions: The complaint's infringement theory focuses heavily on the data capture and analysis functions. A key point of contention will likely be what evidence supports the allegation that the Accused Systems perform the active limitation of "selectively control[ling]" RCLs to position railcars, as required by the claim. The complaint's flowchart of Roboflow's solution does not explicitly show a locomotive control step (Compl. ¶36, p. 8).
’183 Patent Infringement Allegations
| Claim Element (from Independent Claim 16) | Alleged Infringing Functionality | Complaint Citation | Patent Citation |
|---|---|---|---|
| A server for controlling a railway yard... comprising: one or more processors; and one or more non-transitory computer-readable storage mediums | The Accused Systems are alleged to operate on a server with processors and memory to perform the claimed functions. | ¶58 | col. 2:20-22 |
| generate a database associated with the one or more railcars based upon the railyard sensor data | Roboflow's server allegedly organizes extracted container numbers, logos, and classifications into a structured digital database for yard operations. | ¶58 | col. 2:23-25 |
| selectively control the one or more RCLs to position the one or more railcars... based upon the railyard sensor data | The complaint alleges Roboflow's server selectively controls RCLs based on sensor data. | ¶56 | col. 2:26-30 |
| identify each railcar of the one or more railcars based upon the railyard image data | The Accused Systems are alleged to use vehicle-mounted cameras to collect image data and apply OCR and machine learning to identify assets. A provided visual shows a system identifying railroad cars from images (Compl. ¶38, p. 9). | ¶58 | col. 2:5-7 |
Identified Points of Contention
- Scope Questions: Similar to the ’148 Patent, a dispute may arise over whether the accused server, as a standalone component, infringes a claim directed to "controlling a railway yard." This will depend on the server's integration and interaction with other yard components like sensors and RCLs.
- Technical Questions: As with Claim 1 of the ’148 Patent, the allegation that the accused server "selectively controls" RCLs appears to be a critical and potentially contentious point. The evidence cited in the complaint focuses primarily on data processing and asset identification, leaving the specific mechanism of locomotive control less defined.
V. Key Claim Terms for Construction
The Term: "server configured to... selectively control" (’148 Patent, Claim 1) / "selectively control" (’183 Patent, Claim 16)
- Context and Importance: This term is critical because it defines the active, operational step required of the server. The infringement analysis for both lead patents may turn on whether the accused server merely provides data to a human operator or another system, or whether it directly issues commands or otherwise "controls" the physical movement of locomotives.
- Intrinsic Evidence for Interpretation:
- Evidence for a Broader Interpretation: The specification describes the server controlling RCLs "to position the sets of railcars" (’148 Patent, Abstract), which could be argued to encompass providing positioning data that is then acted upon, not necessarily requiring direct command issuance.
- Evidence for a Narrower Interpretation: Figure 3 of the patents explicitly discloses a "LOCO MODULE" for "CONTROL LOCOMOTIVE MOVEMENT" that generates "MOVEMENT INSTRUCTIONS" and a "VTOS HE COMM" module that "BUILDS CONTROL MESSAGES" (’148 Patent, Fig. 3, elements 131, 133). This may support a narrower construction requiring the server to generate and transmit specific, machine-executable control commands.
The Term: "railyard sensors" (’148 Patent, Claim 1)
- Context and Importance: The definition of this term will determine the scope of the physical hardware required by the claim. Practitioners may focus on this term to ascertain whether the standard cameras allegedly used by Defendant's customers fall within the claimed definition.
- Intrinsic Evidence for Interpretation:
- Evidence for a Broader Interpretation: The specification states that the "plurality of railyard sensors may comprise an image sensor... and a proximity sensor" (’148 Patent, col. 2:63-65). The use of "may comprise" suggests that this is an exemplary, non-limiting list, and could be read to include any sensor that generates data about railcars in a yard.
- Evidence for a Narrower Interpretation: The detailed description and figures focus specifically on "PTZ cameras," "fixed cameras," and "3D sensors" as the operative sensors for data collection (’148 Patent, Figs. 2-3). This could support an argument that the term is limited to the types of sensors disclosed as part of the inventive system.
VI. Other Allegations
- Indirect Infringement: The complaint alleges both induced and contributory infringement for all three patents. It claims Defendant encourages and enables its customers to infringe by providing "written instructions, tutorials, source code, training materials, and other technical documentation" (Compl. ¶¶72, 80, 87). It further alleges the Accused Systems are a material part of the patented inventions, are not staple articles of commerce, and are especially adapted for an infringing use (Compl. ¶¶93-94, 100-101, 107-108).
- Willful Infringement: Willfulness is alleged based on Defendant’s purported actual knowledge of the Asserted Patents, established "no later than December 4, 2024, when it received written notice from ConGlobal" regarding the ’148 Patent (Compl. ¶114). The complaint alleges that despite this knowledge, Defendant continued its infringing conduct without taking meaningful steps to avoid infringement (Compl. ¶¶120, 122).
VII. Analyst’s Conclusion: Key Questions for the Case
- A core issue will be one of divided infringement and control: Does Defendant, by providing a software platform and instructional materials to customers who operate the physical hardware (cameras, vehicles, locomotives), exercise sufficient direction or control over its customers' activities to be held liable for infringing system and server claims that require the combined operation of both software and hardware?
- A key evidentiary question will be one of functional performance: What evidence will demonstrate that the accused server performs the specific, active step of "selectively control[ling]" remote control locomotives to position railcars, as required by the claims, or does its functionality end at data analysis and inventory reporting?