Service PawsService Paws
Todos los capítulos

Capítulo 4 · Versión 1.0 · En vigor desde el 4 de julio de 2026

Task Training & Testing Framework

SPTCS · Version 1.0 · Applies to: all disciplines — chapters 5–13 build on this

La norma se redacta y mantiene en inglés, su idioma de referencia.

This chapter defines what qualifies as a task, how tasks must be proofed, and the common rules every discipline task test follows. Passing a discipline's task test records the task testing date (second milestone).


4.1 What qualifies as a task

A certifiable task must be all of the following:

  1. Trained — deliberately built through a documented training process (shaping, capturing, scent imprinting…), not merely an innate behaviour that happens to occur. (Documented innate alert behaviour is recorded separately — §0.5.)
  2. Defined — it has a written description: the trigger (cue or event), the behaviour chain, and the completion criterion.
  3. Repeatable — it can be performed on demand or on a reproducible simulated event, so it can be tested.
  4. Mitigating — it addresses a specific effect of the handler's disability.
  5. Safe — for the dog, the handler, and the public.

Not tasks: presence/comfort alone; "emotional support"; behaviours only the handler can perceive with no observable behaviour chain; any aggression-adjacent behaviour ("blocking" is a positional task and must be entirely passive).

4.2 Task documentation (the task sheet)

Every certified task is written up on a one-page task sheet (template in Chapter 16):

  • Task name and discipline
  • Trigger: cue (verbal/hand) or event (odour, sound, behaviour episode, alarm)
  • Behaviour chain, step by step
  • Completion criterion (what "done" looks like)
  • Proofing environments completed
  • Known failure modes and the retraining plan

The dog's registry record lists its certified tasks by name.

4.3 Proofing requirement

Before assessment, each task must be demonstrated at threshold in all three environment classes (§2.2), and — where the task will realistically be needed — under its real conditions:

  • Night/awake-from-sleep versions for alerts that must wake a handler (diabetic night alert, smoke alarm).
  • Public versions for tasks performed in public (DPT in a store aisle, retrieval in a restaurant).
  • Third-party versions where the task involves other people (summoning a household member).

4.4 Reliability thresholds

CategoryThreshold
Standard tasks≥ 90% success across assessed trials, no environment class below 80%
Safety-critical tasks (marked in discipline chapters — e.g., smoke-alarm alert, emergency summon chain)100% in assessment
Detection discrimination (odour work)Per discipline chapter — typically ≥ 90% hit rate with defined false-alert limits on blanks

Trials must be distributed (across sessions/days), not massed in a single run.

4.5 Simulated events

Alert/response disciplines are tested on staged events (a scripted episode, a planted sample, a played alarm). Rules:

  • The dog must not be able to key off unintended cues: the person staging the event follows a script; sample placements are done by a third party where the protocol requires blind testing.
  • Blind trials: where a chapter requires them, the handler/trainer does not know which presentations are hot vs. blank.
  • Simulations are standardized per discipline chapter so results are comparable between dogs and trainers.

4.6 Evidence & records

  • Video: when Service Paws oversees remotely, task tests are recorded as continuous, unedited video. Even with in-person oversight, video of the formal test is strongly recommended and retained (Chapter 14).
  • Logs: alert disciplines require live-alert logs (dated entries, what the dog did, corroboration — meter reading, witnessed episode, device data). Log standards are defined per chapter.
  • Scoresheets: every formal test uses the Chapter 16 sheet for its discipline.

4.7 Maintenance expectation

Certification reflects tested ability at the time of assessment. Handlers are advised in writing at issuance:

  • Tasks degrade without practice — refresh each certified task at least weekly.
  • Detection dogs need scheduled odour refreshers with verified samples (per discipline chapter).
  • Service Paws may require re-assessment after an incident report or a substantiated performance complaint (Chapter 15).

4.8 Relationship to the three milestones

Task testing may be assessed only after the standards assessment (Chapters 1–2) is passed. The PAT (Chapter 3) may be taken before or after task testing — but certification issues only when all three dates are on record.