# Eco Bag — Design Research Three parallel research streams. Distilled into a single decision document driving the v2 3D rebuild. --- ## STREAM A · Tote bag canon (history + construction) **Origin.** L.L.Bean *Ice Carrier* (1944) — 24-oz oatmeal canvas, double-base, riveted handles, rated ~500 lb. Pulled after one season. Revived 1965 as *Boat & Tote*, two trim colors over natural body. Free monogramming 1968 — first mass "named" bag. **Bonnie Cashin** (Coach, 1962) introduced the brass turnlock and side-tab gussets, making leather totes daywear. **Anya Hindmarch "I'm Not a Plastic Bag"** (2007) — Columbus Circle drop sold 3,000 units in 29 minutes; UK plastic-bag use fell from 10.6bn → 6.1bn over 4 years. **Canonical construction — the spec we should model from:** | Element | Standard | Why it matters for the 3D model | | --- | --- | --- | | Fabric weight | 8–24 oz canvas (Boat & Tote = 24 oz) | Heavy canvas = sharper edges, holds shape | | Panel count | **3-panel** (front + back + bottom gusset) or **5-panel** (+ 2 side gussets) | Visible seam lines define the silhouette | | Strap construction | Self-fabric folded + topstitched, **continuous wrap under base** | Straps must visually pass under the bag | | Strap attachment | **X-stitch box** or industrial **bartack** | Render as visible dense stitching at stress points | | Top hem | Folded 1–2", double-stitched | Slightly thicker top rim | | Hardware | Tradition is **zero**. Modern adds: D-rings, magnetic snaps, YKK zips | Restraint signals quality | | Lining | Premium adds cotton/nylon + 1 slip + 1 zip | Visible at the opening | | Corner | Boxed (sewn diagonal, trimmed) = rigid base; flat = soft drape | Choose per bag intent | **Modern materials (Portuguese sourcing).** Amorim cork (cork oak harvested every 9–12 years), Burel recycled-wool felt (Serra da Estrela), rPET (Patagonia Recycled Market Tote), Piñatex (Ananas Anam, since 2015), MIRUM (Stella McCartney 2023), MycoWorks Reishi (Hermès 2021), apple leather (Frumat), Sea2See ocean nylon. --- ## STREAM B · What makes good design (applied to bags) **Dieter Rams' 10 (Weniger, aber besser)** — every panel, pocket, strap must earn its presence; strip until removing one more breaks the bag; reject trend-coded finishes; mono-material where possible. **Don Norman — affordances + signifiers.** Handle should *look* graspable (slight wrap, contrasting material). Opening should *signal* itself (zipper pulls sized for fingers, magnetic flap that lifts at rest). Conversion mechanism should *expose* itself (visible G-hooks, color-coded webbing endpoints). **Maeda — Laws of Simplicity for convertibles.** - **Reduce** — 2–3 modes max, each justified - **Organize** — all conversion hardware on one panel, hand finds next step within reach of last - **Time** — the conversion moment must feel *fast and rewarding* (like a well-tuned car door) **Antifragile materials (Taleb).** Waxed canvas, vegetable-tanned leather, raw selvedge denim, solid brass — patina-coded; PU-coated nylon, chrome-plated zamak, printed graphics — fragile. **Convertibility = clarity of state (Nielsen).** The two modes must be visually distinct from across the room. Hidden-strap bags that look identical in both modes confuse the wearer and observers. **Material honesty (Vitsoe).** Show what it is. Reject "vegan leather" that mimics grain stamps. Solid brass exposed, not painted to fake another metal. **The 80/20 of carry experience:** 1. Weight distribution (load near spine, high for backpacks; short drop + structured base for totes) 2. Strap drop length (~55cm cross-body; ~22–25cm tote handle clearance) 3. One-handed opening accessibility 4. Weather resistance (coated bottom, taped seams, storm flap) **Anti-patterns we must avoid.** Fake hardware (decorative buckles that don't buckle). Oversized branding. Novelty mechanisms with no daily benefit. Over-engineered 6-in-1 convertibility nobody uses. Decorative pockets too small for anything. Tactical-cosplay MOLLE. **All failures share one root: features designed for the showroom photo, not the 1,000th day of carry.** --- ## STREAM C · The 10 concepts — engineering verdict Researched against patents, brand precedents, manufacturability. Three categories: ### PROVEN (1 concept) — ships today, just needs to be made - **#1 CUBE-FOLD** — Hidden-strap convertibles are a mature category. Knack Pack Series 2, Bellroy Transit, Boundary Errant. **Engineering fix: drop the slider rails.** Every shipping convertible uses zip-and-tuck — rails add 80–150g, jam under load, fail lateral-shear. Replace with YKK #8 reverse-coil zip around back-panel perimeter. Ship in 6 months. ### PLAUSIBLE (4 concepts) — buildable with named engineering fixes - **#3 PETAL BUCKET** — Furoshiki *tesage bukuro* precedent. Closest commercial: Link Collective Arcs. **Engineering fix:** use Fidlock rotary buckle (mature hardware); pre-reinforce 2 "strap" petals with EVA + webbing spine; others stay decorative. **Risk:** 80→20mm trapezoidal strap = sash-bag comfort, max 5kg load. - **#5 RIDGE-SPINE** — Peak Design Everyday + Boundary Errant lineage. **Engineering fix:** ridge **must fatten to 45mm** (current 25mm is geometrically impossible for ergonomic 8–12mm padded straps). Use YKK Aquaseal #5 curved zip with TPU bonding. Risk: spine-zip aesthetic vs ergonomic strap padding. - **#6 TAPERED ZIGGURAT** — Aer Travel Pack + Tortuga lineage for tier stacking, but no one does the 180° back-flip hinge. **Engineering fix:** HDPE living hinge with TPU laminate (custom tool, $8k); or metal piano hinge (+150g, heavy). Empty weight estimate 1.6–1.9kg. - **#7 HORN-LOOP ROLL-TOP** — Topo Klettersack roll-top + retractable strap patent US 5,529,229A. **Engineering fix:** license the spring-reel + add separate cam-lock (2-step user flow). Ripcord-through-fabric-friction sags under sustained load >3kg without lock. ### EXPERIMENTAL (5 concepts) — research projects, multi-year development - **#2 HELIX TOTE** — geometry that makes stable tote (wraps locked to base) is **incompatible** with geometry that makes shoulder loops (wraps must release). Resolving needs cam-locks at every turn intersection — kills the "single band" elegance. No commercial precedent. - **#4 SADDLE TOTE** — single-panel saddle has **no closed volume**. Hypar requires 4 boundary anchors at alternating heights to self-tension. Adding pockets destroys the minimal-surface premise. **This is sculpture, not a bag.** - **#8 INFLATABLE EXOSKELETON** — Klymit ships flat air-frame back panels (US 8,215,526); 3D Voronoi welding isn't manufacturable at scale today. 0.2 bar spec is 4–40× higher than architectural inflatables. Single 1mm puncture eliminates strap function — backup carry mode mandatory. - **#9 MÖBIUS CARRY** — Möbius topology has **one continuous edge**, not two — "both edges as straps" is physically the same edge under each arm twice. Half-twist is invisible at body scale. A cylindrical band (no twist) is more comfortable, equally convertible, visually identical when worn. **Topology lesson, not a wearable improvement.** Real capacity ~2–4L, not 8L. - **#10 MIURA-FOLD ORIGAMI** — JAXA-deployed solar array works because it's frame-held after deployment. Deployed Miura panel is **floppy perpendicular to its plane** — doesn't form a closed 14L volume on its own. Issey Miyake Bao Bao (rigid PVC tiles on mesh) is actually a more realistic path than true Miura-fold. --- ## How this changes the 3D rebuild Each model now needs to express its engineering verdict visually: - **PROVEN** (CUBE-FOLD) → render as a finished commercial product. Visible panel seams, bartacks at strap mounts, proper canvas-weave texture, YKK zip-and-tuck back panel (no fictional rails). - **PLAUSIBLE** (PETAL, RIDGE-SPINE, ZIGGURAT, HORN-LOOP) → render with the named engineering fix applied — fatten the ridge, show the Fidlock buckle, expose the living hinge, draw the cam-lock. - **EXPERIMENTAL** (HELIX, SADDLE, INFLATABLE, MÖBIUS, MIURA) → render as research prototypes. Exposed mechanisms, visible "this is a prototype" character, slightly less finished surface treatment. Keep them but signal the maturity gap. **Material upgrades across all 10:** - Procedural canvas-weave normal map (so cotton/linen/waxed-canvas actually read as woven) - Cork stipple normal map (so cork panels show their cellular surface) - Wool felt micro-noise normal map - Real shadow casting onto the ground plane (DirectionalLight.shadow) - Visible front/back/bottom/gusset seams as construction lines, not just edge stitching - Bartack X-stitch boxes at every strap attachment point - Proper strap drop lengths per the 80/20 rules **Sources for further reading:** L.L.Bean Boat & Tote history, Bonnie Cashin Coach archive, Anya Hindmarch V&A collection, Knack Series 2 specs, Bellroy Tokyo Totepack, Patagonia Recycled Market Tote, MycoWorks Reishi (Hermès), Sea2See ocean plastic, US 5,529,229A retractable strap, US 8,215,526 air-cushion backpack, Klymit Air Beam, Issey Miyake A-POC + Bao Bao, Miura fold (JAXA solar array).