# Design Possibilities: Innovation Within E2510 **Purpose:** Map which QuickChat innovations survive funded configuration compliance — and which need to live elsewhere. Thread the needle between "clinically sterile like every competitor" and "too innovative to fund." **Date:** April 9, 2026 --- ## The Frame E2510 compliance doesn't require us to build another boring grid app. It requires that the funded configuration be dedicated to **speech generation.** That's actually a large design space — much larger than what competitors use. They've all retreated to grids not because grids are required, but because grids are safe and nobody's challenged the assumption. The question for each QuickChat feature is: **"Is this speech generation, or is this something else?"** --- ## Tier 1: Fundable As-Is (Speech Generation, Full Stop) These features ARE the speech generating function. They belong on the funded configuration. No unlock required. ### Scene-Based Navigation (DP-2) Visual Scene Displays are a **clinically validated communication method** with published research (Wilkinson & Light). The funded config can and should use scenes instead of grids. This alone differentiates QuickChat from every funded competitor while being fully compliant. **Why it survives:** A scene with tappable objects that produce speech IS an SGD interface. It's just a better one. ### Speak-Choose-Speak Conversation Loop (DP-6) The core interaction pattern — child speaks, app offers contextual follow-ups, child selects, new options appear. This is **message formulation** — the exact thing E2510 covers. **Why it survives:** Multiple methods of message formulation is literally in the E2510 definition. A conversation loop that generates contextual speech options is the most sophisticated message formulation method on any SGD. ### Fitzgerald Key Color Coding (DP-4) Color-coded grammar categories. Already used by SLPs, already present in competitors' apps. **Why it survives:** It's a visual grammar system that supports communication. No risk. ### Multisensory Feedback on Tap (DP-8) Symbol highlight + spoken word + haptic response on every tap. This is **sensory confirmation that speech was generated** — the same principle as the audio click on a physical SGD button, extended to multiple modalities. **Why it survives:** Every SGD provides feedback when speech is generated. QuickChat just does it in three channels instead of one. ### Positions Are Promises / Motor Planning (DP-1) Spatial consistency for motor learning. LAMP's entire methodology is built on this, and LAMP is one of the most-funded AAC apps in history. **Why it survives:** Motor planning support is a clinical feature, not an entertainment feature. LAMP proves this passes PDAC. ### Template Engine Sentence Building (BP-3) Grammar-aware slot filling that generates thousands of utterances. This is the speech generation mechanism itself. **Why it survives:** It's the engine. Without it, there's no SGD. ### All 8 Pragmatic Functions (OP-4) Supporting commenting, protesting, greeting, questioning — not just requesting. Every one of these produces speech output. **Why it survives:** All pragmatic functions are speech. An SGD that only supports requesting is a bad SGD, not a more compliant one. ### Age-Adaptive Complexity (DP-3) Showing 1-word through 4-word options simultaneously. The child self-selects complexity level. **Why it survives:** This is scaffolded message formulation. It's a clinical approach to language development through the SGD. --- ## Tier 2: Fundable With Careful Framing (Clinical Support for Speech) These features support speech generation but need deliberate clinical framing in PDAC documentation and marketing materials. They're defensible but require intentional positioning. ### Synesthetic Grammar Identity (DP-9) Category-specific animation style + sound character on tap (verbs pulse forward, nouns bounce, etc.). This extends Colourful Semantics — a published, evidence-based intervention — from visual to multisensory. **Design possibility:** Keep animations subtle and synchronized with speech output. Frame as "multisensory grammar cueing" in documentation. Cite Colourful Semantics research (Bryan, 1997; Bolderson et al., 2011 — 12-18 month language gains). **Risk if overdone:** Flashy animations that look decorative rather than clinical. Keep them functional — they should feel like feedback, not entertainment. ### Feelings Expression Tapping a feeling face that produces speech ("I feel sad") is speech generation. No question. **Design possibility:** Feelings as direct speech output (Tier 1). Where it gets into Tier 2: if there's an expanded "check-in" flow with multiple steps, visual journaling, or mood tracking beyond producing speech, that starts looking like wellness rather than SGD. **Threading the needle:** Keep the funded config's feelings feature as simple speech output. The richer emotional exploration flow can live in Tier 3. ### Parent/Caregiver Coaching (OP-5) In-app guidance teaching aided language stimulation, modeling techniques, wait time. **Design possibility:** Frame as "communication partner training" — directly supporting the SGD's therapeutic use. SLPs already prescribe partner training as part of AAC intervention. **Risk:** If it looks like an educational app teaching parenting skills rather than supporting device use, it drifts. Keep it tightly coupled to "how to use this device with your child." ### Data Logging / Progress Tracking Usage data, communication attempts, vocabulary growth metrics for SLP review. **Design possibility:** Frame as "clinical outcome measurement." SLPs need this data for IEP goals and continued funding justification. It actually SUPPORTS E2510 compliance by demonstrating the device is being used for speech. **No risk.** This is a clinical tool. Competitors already include it (Avaz does, Proloquo2Go does). Build it. --- ## Tier 3: Post-Funding Unlock Only (Innovation That Can't Live on Funded Config) These features are the engagement innovations that differentiate QuickChat — but they can't be on the funded configuration. They need the PRC-Saltillo treatment: available after delivery, behind a toggle or unlock. ### Rewards, Celebrations, Points, Streaks Motivational systems that reward communication attempts. **Why it can't be funded:** Looks recreational. No competitor includes anything like this on a funded config. PDAC reviewers would flag it. Even if defensible in theory ("reinforcement for communication"), the visual impression is "game," and PDAC impressions matter. **Design possibility:** Build a "Clinician Mode" vs. "Full Experience" toggle. Clinician Mode = funded config (no rewards visible). Full Experience = everything on. Default to Clinician Mode on QuickTalker Freestyle, Full Experience on consumer installs. ### Gamified Sentence Builder Drag-and-drop sentence construction as play. **Why it can't be funded:** The word "game" is in the concept name. Even though the underlying function (building sentences) is speech generation, the play framing makes it look recreational. **Design possibility:** Reframe the funded-config version as "Structured Sentence Practice" — same interaction, stripped of game elements (no points, no scoring, no level progression). Keep the core interaction (dragging words into sentence slots) because that's message formulation. Add the game layer in Tier 3 unlock. ### AR Word Explorer Point iPad at objects, tap to hear words. **Why it can't be funded:** This looks like an educational AR app, not an SGD. Even though it generates speech, the primary impression is "exploration tool," not "communication device." **Design possibility:** Keep it as a consumer/App Store differentiator. This is one of QuickChat's most innovative ideas — it just can't lead the funded pitch. ### ShapeSpeak (Speech Production Game) Mouth shape visualization game for verbal-trajectory kids. **Why it can't be funded:** It's explicitly a game, and it targets speech production (articulation), not speech generation (AAC output). E2510 covers devices that generate speech FOR the user, not devices that teach users to produce their own speech. Different clinical function entirely. **Design possibility:** Consumer-only feature. Could also be positioned as a separate app or in-app purchase. ### Play Together Mode / Peer Communication Split-screen interaction where two children share the device. **Why it needs careful handling:** The communication itself (both children producing speech via the device) is fundable. The "play" framing and split-screen interaction could look like a social game. **Design possibility:** On funded config, frame as "Shared Communication" — two users accessing the SGD simultaneously. Strip the play language. In consumer mode, bring back the full Play Together framing. --- ## The Architecture Recommendation ### Two modes, one app: **Meadow Clinical** (funded configuration) - Launches as default on QuickTalker Freestyle - Scene-based navigation (Tier 1) - Speak-choose-speak loop (Tier 1) - Fitzgerald Key + multisensory grammar cueing (Tier 1-2) - All 8 pragmatic functions (Tier 1) - Age-adaptive complexity (Tier 1) - Data logging for SLP (Tier 2) - Communication partner guidance (Tier 2) - No rewards, no games, no AR **Meadow Full** (consumer + post-funding unlock) - Everything in Clinical, plus: - Rewards and celebrations - Gamified sentence builder - AR Word Explorer - Play Together Mode - ShapeSpeak - Richer emotional exploration - All the features that make kids actually want to use the app ### Why this works: - The funded config is **already more innovative than any competitor** — scenes instead of grids, conversation loops, multisensory grammar, age-adaptive scaffolding - The unlock adds engagement features that address the 60%+ abandonment problem - Follows the PRC-Saltillo precedent that the industry and PDAC already accept - AbleNet maintains E2510 billing capability on the Freestyle - QuickChat has a consumer App Store story that no locked-down competitor can match ### What to present to AbleNet: Don't present this as "we'll strip out the fun stuff." Present it as: "The funded configuration is already the most innovative SGD software on the market — scenes, conversation loops, multisensory grammar. No competitor has any of this on a funded device. AND we have an engagement layer that addresses the abandonment crisis, available for families post-funding. We're not choosing between innovation and compliance. We're doing both." --- ## Open Questions for SLP Consultation 1. Would an SLP reviewing the funded config see speak-choose-speak as a clinical interaction or a chatbot? (We think clinical — but confirm) 2. Are synesthetic grammar animations (DP-9) standard enough to survive PDAC, or do they need published evidence cited in the submission? 3. Does communication partner coaching belong on the funded config or is it scope creep? 4. How do SLPs currently document QuickTalker Freestyle use for continued funding? What data would they want from our app? 5. Has AbleNet's Freestyle ever had a funding denial related to software features?