Task and Capability Analysis¶
Analysis method — instantiates Affordance Shaping
Decomposes the goal into the actions it requires and checks each against what the agent can actually perceive, reach, and do — locating where the task outruns the agent's capability.
Before you can make the right action easy, you have to know what the right action demands and whether the agent can meet it. Task and Capability Analysis decomposes the goal into the sequence of sub-tasks and operations that reach it, and holds each one against an explicit model of the agent — its perception, reach, dexterity, memory, attention, prior knowledge. Its defining move is the match: it is the mechanism that names, step by step, where the task's demand exceeds the agent's capability, and it is also where the goal path is separated from the harm path so the rest of the archetype knows which way to steer. It produces no cue, no default, and no test — it produces the specification the shaping later fills.
Example¶
A device maker is designing an insulin pen for people to inject themselves at home, many of them elderly. Rather than assume "the user injects a dose," the analysis breaks the goal into its real operations: read the dose from a small dial, grip and hold a steady position, prime the needle, push with enough sustained force, count the hold time, dispose safely. Against each step it lays a capability model of the actual population — reduced grip strength, presbyopia, tremor, occasional confusion about units.
The mismatch jumps out: the dial numerals are below the reading capability of a chunk of the intended users, and the push force at the end of travel exceeds a weak thumb. Just as important, the analysis marks the harm path — dialing units as though they were tenths — as a step one confusion away from the goal path. None of this is fixed here; the output is a map of where capability falls short and where the dangerous branch sits, which is what tells the signifier, default, and testing mechanisms exactly what to shape.
How it works¶
- Decompose the goal into operations. The task is broken into nested sub-goals and the concrete actions that satisfy them, in the order and conditions they occur — the hierarchical-task-analysis move.[1]
- Model the agent, not an idealized one. Perception, reach, strength, dexterity, working memory, attention, and prior knowledge are stated as a capability profile of the real intended agent, including its variation.
- Match demand to capability, step by step. Each operation's demand is checked against the profile; the gaps — where demand outruns capability — are the output that matters.
- Separate goal path from harm path. The intended route to the goal and the nearby error or harm routes are named explicitly, so later mechanisms can invite one and discourage the other.
Tuning parameters¶
- Decomposition depth — how finely the goal is broken down. Deeper analysis catches subtle capability gaps but costs effort and can drown the signal in detail.
- Capability breadth — how many dimensions of the agent are modeled (motor, sensory, cognitive, motivational). Wider profiles catch more mismatches; narrow ones move faster.
- Population span — a single "typical" agent or the full distribution of intended agents. Modeling the tails is where accessibility problems surface.
- Normative vs. observed — whether the task is analyzed as it is supposed to be done or as people actually do it. The second is truer and requires observation to feed it.
When it helps, and when it misleads¶
Its strength is that it locates the mismatch before anything is built, and it does so in the agent's own terms — so the eventual affordance is aimed at a real gap rather than a guessed one. It is the upstream diagnosis the whole archetype leans on: signifiers, defaults, and friction all need to know which step is hard and which branch is dangerous.
Its failure modes are those of any armchair analysis. It can model the idealized task instead of the messy real one, and it can collapse a diverse population into a fictional "average user" who has everyone's abilities and none of their limits — the flaw that lets an inaccessible design pass on paper. The classic misuse is running the analysis after the design is fixed, to rationalize choices already made rather than to find their gaps. The discipline that guards against it is to ground the task model in real observation and to model the capability distribution, tails included, then hand the result to a test that can contradict it.
How it implements the components¶
Task and Capability Analysis fills the diagnostic, specification side of the archetype — the parts that define what must be shaped, not the shaping itself:
agent_capability_model— its core artifact: an explicit profile of what the intended agent can perceive, reach, remember, and do, with its variation.goal_and_harm_path_definition— the decomposition names the intended route to the goal and the adjacent error/harm routes, giving the archetype its target and its hazard.
It stops at specification. The physical feasible-action map is Robot Action-Space Mapping; the perceivable cues are Signifier Prototyping; and whether the capability model was right is decided by Usability or Field Test, which can contradict it with real behaviour.
Related¶
- Instantiates: Affordance Shaping — it supplies the demand-versus-capability specification the rest of the shaping is aimed at.
- Sibling mechanisms: Usability or Field Test · Signifier Prototyping · Robot Action-Space Mapping · Safe Default or Preselected Path · Wayfinding Marker · Contextual Inquiry or Walkthrough · Affordance Audit
Editorial Notes¶
Form Classification¶
Form family: Analysis, Modeling & Optimization
Rationale: Task and Capability Analysis operates as an analytical, modeling, inference, comparison, or optimization procedure that derives insight or a solution because it decomposes the goal into the actions it requires and checks each against what the agent can actually perceive, reach, and do — locating where the task outruns the agent's capability.
Independent corroboration: The frozen evidence defines Task and Capability Analysis as 'Decomposes the goal into the actions it requires and checks each against what the agent can actually perceive, reach, and do — locating where the task outruns the agent's capability', so its operative form is Analysis, Modeling & Optimization.
Nearest alternative: Representation, Specification & Plan — Task and Capability Analysis includes features of a static representation, map, specification, schema, or prospective plan that externalizes information, but its defining operation is an analytical, modeling, inference, comparison, or optimization procedure that derives insight or a solution.
Review outcome: Independent reviewer agreement; medium confidence.
Origin Attribution¶
Primary origin: Human-Computer Interaction
Origin pattern: Single lineage
Present-day reach: Multi-domain
Rationale: The defining operation is: Decomposes the goal into the actions it requires and checks each against what the agent can actually perceive, reach, and do — locating where the task outruns the agent's capability. In the human_computer_interaction lineage, that operation is specifically evidenced by authoritative or primary work that decomposes mission tasks and evaluates human perception, action, workload, and capability against task demands. This makes human_computer_interaction the best historical origin, while the retained alternates document contributing methods and later applications rather than being mistaken for coequal origins.
Related originating lineages:
- Computer Science & Software Engineering — Computer science and software-engineering practice supplies a parallel or contributing lineage for the mechanism's defining operation: decomposes the goal into the actions it requires and checks each against what the agent can actually perceive, reach, and do — locating where the task outruns the agent's capability.
- Organizational & Management Science — organizational_management supplies a historically relevant parallel or contributing practice for the defining operation—Decomposes the goal into the actions it requires and checks each against what the agent can actually perceive, reach, and do — locating where the task outruns the agent's capability—but the evidence does not make it the best primary lineage.
- Psychology — Experimental, clinical, and behavioral psychology supplies a parallel or contributing lineage for the mechanism's defining operation: decomposes the goal into the actions it requires and checks each against what the agent can actually perceive, reach, and do — locating where the task outruns the agent's capability.
- Robotics & Automation — Robotics, sensor fusion, and automation supplies a parallel or contributing lineage for the mechanism's defining operation: decomposes the goal into the actions it requires and checks each against what the agent can actually perceive, reach, and do — locating where the task outruns the agent's capability.
Review resolution: The blind reviewers disagree on primary lineage (organizational_management versus human_computer_interaction), so I adjudicated the mechanism rather than inheriting either label. The defining operation is: Decomposes the goal into the actions it requires and checks each against what the agent can actually perceive, reach, and do — locating where the task outruns the agent's capability. In the human_computer_interaction lineage, that operation is specifically evidenced by authoritative or primary work that decomposes mission tasks and evaluates human perception, action, workload, and capability against task demands. This makes human_computer_interaction the best historical origin, while the retained alternates document contributing methods and later applications rather than being mistaken for coequal origins. The cited NASA Human-Systems Integration Handbook directly supports the mechanism-specific operation and its disciplinary lineage. I retain all independently explained historical alternates without a numeric cap. origin_mode=single_lineage records how the mechanism arose; domain_reach=multi_domain separately records how broadly it can now be applied.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Researched adjudication after independent review; high confidence.
Sources consulted:
Notes¶
This mechanism predicts; it does not validate. A capability profile is a hypothesis about the agent that can be wrong, and the analysis is only as good as the observation feeding it — so it is best treated as the front half of a loop that closes with Usability or Field Test, not as a verdict on its own.
References¶
[1] Hierarchical Task Analysis decomposes a goal into nested sub-goals and operations, with plans describing when each is done; a long-standing ergonomics and HCI method (Annett & Duncan, 1967). It is the standard way to expose, operation by operation, where a task's demands sit relative to an agent's capabilities. withdrawn registry ↩