This dataset contains 12 patent families published between 2017 and mid-2026, drawn from a search focused on computation offloading, task partitioning and orchestration under the G06F9, H04L67 and H04W28 classifications. That is a small corpus for a field this broadly named, which tells you the search terms are catching a specific technical slice rather than the whole edge-computing space. Broader searches on adjacent terms like network function virtualisation or general edge orchestration would likely surface additional related families outside this scope.

Intel Corporation holds the most-cited lineage in this dataset, with a family of application computation offloading patents that includes a record cited 496 times, alongside related publications cited 80, 51, 21 and 9 times respectively. These are all variants of the same titled invention progressing through publication and grant stages, which means the underlying claim territory has been under continuous refinement rather than being a single one-off filing. Any competitor targeting mobile-to-edge offloading with energy and latency budgeting should review this lineage first.

Filing activity is still accelerating based on the data available: annual filings moved from 2 in 2017 to a peak of 3 in 2023, with a low point of 1 at the 2022 midpoint. The 2026 figure of 2 looks lower only because publication typically lags filing by around 18 months, so recent filings have not fully surfaced yet. Taken together, the trend does not show a field that has peaked and is now declining.

AI-driven offloading decision logic, classified under G06N, appears in only 1 of the 12 records in this dataset, despite the other core classifications (G06F, H04L, H04W) each appearing in around 10 of the 12. That gap suggests claim space around machine-learning-based task scheduling or predictive offloading decisions is comparatively open. Multi-RAT-aware scheduling and data-locality-based partitioning also show thin representation relative to the breadth of the search string.

US11695821B2 covers offloading of computationally intensive tasks between devices with explicit consideration of energy consumption and latency budgets, including scenarios exploiting multiple radio access technologies. It does not block all offloading work outright, but new filers should check their claims against this family and its earlier-cited relatives, particularly if the invention touches RAT-aware scheduling or joint energy-latency budgeting. Claims narrower in scope, such as those focused purely on data locality or AI-driven decision models, are more likely to sit outside its coverage.