search.noResults

search.searching

saml.title
dataCollection.invalidEmail
note.createNoteMessage

search.noResults

search.searching

orderForm.title

orderForm.productCode
orderForm.description
orderForm.quantity
orderForm.itemPrice
orderForm.price
orderForm.totalPrice
orderForm.deliveryDetails.billingAddress
orderForm.deliveryDetails.deliveryAddress
orderForm.noItems
Column: Silicon systems design


Figure 1: Controlled engineering context for a verification agent


a formal environment and remove the behaviour in which the bug exists. Te challenge is not tool execution,


but preserving the connection between engineering intent, verification activity and defensible evidence.


The context an agent must understand A useful verification agent needs more than RTL and a general-purpose language model. It may require specifications, verification objectives, testbench structure, assertions, assumptions, coverage models, reference models, waivers and sign-off criteria. Tis context must be current and


trustworthy. If the specification is ambiguous or the verification plan is outdated, an agent can act consistently on the wrong information.


Context must therefore become


a controlled engineering asset, with ownership, versioning and traceability. Agentic AI may navigate verification environments, but it still depends on established methodologies such as UVM and Portable Stimulus.


From tool automation to workflow orchestration Te largest opportunity may lie not within a single task, but between tasks. Requirements become plans, plans


become tests, properties and coverage models. Regression results move into debug, which then creates design changes, new tests and revised assumptions. Each hand-off introduces delay and the risk of losing context. An agentic workflow could maintain these relationships. An agent investigating


a failure could retrieve the requirement, identify the relevant test and assertion, compare revisions, propose the next experiment and update the investigation record. Success should be measured not by


completed actions, but by earlier defect discovery, improved traceability and stronger closure decisions.


Human oversight must be designed into the flow A controlled verification agent needs a defined objective, clear authority boundaries and traceable outputs. Engineers should be able to identify the specification, design revision, model configuration, tool version and input data associated with each action. The agent must distinguish


between a confirmed tool result, an www.electronicsworld.co.uk September 2026 13


Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24  |  Page 25  |  Page 26  |  Page 27  |  Page 28  |  Page 29  |  Page 30  |  Page 31  |  Page 32  |  Page 33  |  Page 34  |  Page 35  |  Page 36  |  Page 37  |  Page 38  |  Page 39  |  Page 40  |  Page 41  |  Page 42  |  Page 43  |  Page 44