# LrrK Kinematic Consumer Risk Bulletin **LRRK Watch / Open-Source Intelligence** — Consumer systems that sense, move, and act in homes and yards. **Issue type.** Monthly filing. | Field | Value | | --- | --- | | Collection cutoff | 30 April 2018 | | Platform | Quiet Systems | | Window | 1–30 April 2018 | | Data label | Open-source intelligence // LRRK analytic product // retrospective monthly filing | | Handling | Public-source material. Image rights require separate verification before publication. Retrospective reconstruction from contemporaneous public records. | | Scope | Consumer kinematic electronics only (home, yard, companion, educational). | | Classes in frame | Robot vacuums / mop combos · robotic lawn mowers · window-cleaning robots · pool-cleaning robots · robot pets / companion animals · social / desk companions · educational / programmable kits · robotic litter boxes · security / home patrol robots · consumer robotic arms · elder-care / assistive companions · kitchen / food-prep robots · toy / entertainment robots · niche outdoor / garden robots · emerging general-purpose or humanoid home robots | | Spine | CPSC, Health Canada, EU Safety Gate, UK OPSS, manufacturer recall pages, named newsrooms, dockets. | | Color | Message boards, consumer forums, and Reddit — annotated context only. Never Fact without a reputable primary. | ## Executive read ### BLUF Stiftung Warentest’s robotic-mower test lands before the season. Eight machines; none better than *befriedigend*; two *mangelhaft* after driving onto a crawling-child foot dummy and cutting the shoe. Standing children were detected; a lying child’s outstretched hands were not. This is a lab result, not a named injury and not a recall. ### Key judgments 1. **[Assessment — High confidence]** The 26 April trade copy, sourced to Stiftung Warentest and matching the May *test* Heft, is an in-window child-vs-blade control finding on current SKUs. 2. **[Assessment — High confidence]** Dummy cuts are not a hospital count. No official recall cites this test this month. 3. **[Inference — Moderate confidence]** A1:2018’s standing-child probe and Warentest’s crawling/lying dummies are not the same test. The gap the hospital series will later name is already visible in the dummy. ### Analytic labels - **Fact** — verified reporting or primary record - **Assessment** — analytic judgment - **Inference** — reasoned but not directly observed - **Uncertainty** — unresolved information gap ### Source grades - **A1** — authoritative primary record / directly confirmed - **B1–B2** — generally reliable and corroborated - **C3** — useful but requires caution (typical for forums) --- ## 01. Stiftung Warentest — eight robotic mowers; two cut the crawling-child shoe *Event / publication dates: 26 April 2018 (Gabot, Quelle: Stiftung Warentest); May 2018 *test* Heft as reported by named trade press* **Product class.** Robotic lawn mowers **Verified record — [Fact — B1]** Gabot, 26 April 2018, citing Stiftung Warentest: eight robotic mowers tested (five for lawns ≤1,000 m²; three for larger; 800–2,850 EUR). None scored better than *befriedigend*. Two scored *mangelhaft* on safety. Testers simulated a standing child, a crawling child, and a lying child. “Ein stehendes Kind erkannten alle Rasenroboter. Bei krabbelnden Kindern hielten sechs ebenfalls rechtzeitig an. Zwei fuhren auf die Fußattrappe auf und hinterließen deutliche Schnitte an dem Kinderschuh. Barfuß hätte das zu heftigen Verletzungen geführt.” None of the eight stopped for the outstretched hands of a lying child; “Die Messer erreichten jedes Mal die Prüffinger.” [Gabot, 26 April 2018](https://www.gabot.de/mehr/dossiers/baumschultechnik-2019/news-baumschultechnik-2019/rasenroboter-keiner-ohne-unfallrisiko-391174.html). May Heft confirmed by named trade press (Baumarktmanager; housecontrollers 14 May). Official test.de HTML for the 2018 table was not independently opened on this pass. **This is not a named human injury and not a regulator file.** The two *mangelhaft* brand/model names are not stated on the opened Gabot page and are not invented here. A1:2018 is January background. **Threat landscape.** A crawling or lying child is below the obstacle profile a shock-sensor mower treats as lawn irregularity. The physical consequence named by the testers is cut injuries on a dummy foot and dummy fingers. **Forum / community notes.** - none independently opened. **Intelligence assessment.** [Assessment — High confidence] Dated in-window lab finding on child-shaped obstacles. [Inference — Moderate confidence] A standing-child probe will not close the crawling/lying case the testers failed. [Uncertainty] Which two SKUs failed; exact *test* table scores without the opened Heft. **Opportunity.** A brand-named pediatric case, or a RAPEX/OPSS alert that cites A1. **LRRK relevance.** Sense / Move / Act. Lab: crawling-child and lying-hand probes, not only standing-foot. Watch: EN 50636 series. **Confidence.** High on the Gabot/Warentest mechanism sentences. Moderate on Heft pagination. Low on unnamed failing models. --- ## Forward indicators 1. Any first named EU child-vs-blade *incident* (this month is a dummy). 2. Dated Jibo / Kuri distress. 3. Any 2018 official consumer-robot recall. 4. aibo US SKU. 5. Brand names on the two *mangelhaft* machines, if a primary states them. > **Collection integrity.** One signal. Official test.de 2018 HTML not opened; Gabot is contemporaneous named trade press quoting Stiftung Warentest. **Excluded:** CPSC 20-115 (2020); ride-on / walk-behind files; Honda Miimo 2024. No April CPSC/Health Canada/OPSS/Safety Gate consumer-robot recall found. Forum empty. *LRRK — security assurance for systems that sense, move, and act.*

L-C3