Week 6 Session: The Vendor Demo Tribunal

Cross-examining three fictional wildlife-safety specification cards — classifying their real function, calculating their real performance, and deciding what evidence would actually justify a recommendation.

29 March 2027

Before the session#

Read this week’s lecture, especially the four-function taxonomy and the three specification cards — you’ll classify and calculate from them directly. Bring a calculator. There’s no submission due this week: the literature review and hotspot analysis isn’t due until 12 April 2027, well after this session.

The specification cards#

The same three fictional systems and the roo bar from the lecture, restated here for quick reference during the session.

The trial table#

A synthetic trial result for one fictional detection system, tested on 100 staged approaches:

Every rate you calculate from this table needs its own denominator — the 40 animal-present trials, or the 60 no-animal trials — never the combined 100.

Lighting, speed and obstruction scenarios#

Four short scenarios for the specification-comparison step:

• Scenario 1: clear daylight, kangaroo stationary at the verge, 50 km/h.
• Scenario 2: dusk, kangaroo partially obscured by roadside scrub, 80 km/h.
• Scenario 3: full night, kangaroo moving across the road, 100 km/h.
• Scenario 4: clear daylight, kangaroo stationary and unobstructed, 90 km/h.

In the session#

This runs as one 50-minute sequence, after a short lecture recap. Work in the same pairs as previous weeks.

1. 5 minutes — pick a claim. Read the lecture’s opening vendor claim again (“our system detects wildlife before you do”). Each pair writes one sentence naming which of the four specification cards they’d initially find most convincing, and why.
2. 10 minutes — classify. For each of the three systems and the roo bar, write down which of the four functions (detection, warning, automatic braking, impact protection) it actually performs. Some cards perform more than one; the roo bar performs none of the first three.
3. 10 minutes — compare against scenarios. For each of the four lighting/speed/obstruction scenarios above, decide which specification cards’ stated operating conditions actually cover it, and which don’t. Write one sentence per scenario naming the gap, if there is one.
4. 10 minutes — calculate. From the trial table, calculate the detection rate, false negative rate and false positive rate, stating the denominator you used for each.
5. 10 minutes — propose a test, or say the evidence is insufficient. For one specification card of your choice, either propose a fair test that would extend its stated evidence to a condition it doesn’t yet cover, or conclude that the evidence supplied is insufficient to recommend it for that condition. Both are acceptable answers — you are not required to pick a winner if the cards don’t support one.
6. 5 minutes — recommend. Write a short recommendation for one system, stated with explicit conditions and caveats (for example: “recommended for daylight, low-speed rural roads; insufficient evidence for dusk or highway speeds”).

Afterwards#

This session’s output is the Brochure Cross-Examination — a one-page pair submission. It’s formative and ungraded: it’s practice at reading a specification honestly and calculating from a trial table, not a mark on which system you preferred. It does not introduce any new assessed requirement of its own.

The Brochure Cross-Examination

• The claim you initially found most convincing (from step 1).
• The actual function of the system behind that claim (from step 2).
• Two operating limits of that system, stated in your own words.
• One calculated performance measure from the trial table, with its denominator stated explicitly.
• One proposed test, or a stated conclusion of insufficient evidence (from step 5).
• Your conditional recommendation from step 6.

Keep your cross-examination. The same habits — checking a denominator, separating a sensor’s claim from a system’s outcome, naming what a specification doesn’t cover — are exactly what the literature review and hotspot analysis rewards when applied to published research instead of a specification card. That assignment is due 12 April 2027 and worth 40% of the unit; nothing from weeks 4 to 6 changes that.