📊 Barashi Defect Dashboard v2

Toyota ASSB · Jan–Jul 2026 · 0 defects · Cleaned Part/Category/Location extraction
BotBot analysis · 2026-07-09
TAGG/T-A (Production self-declared)
QG/IR-QG (Quality Gate)
HANE (Offline repair)

🥧 Parts Distribution (Top 10)

📊 Categories

📊 Top 12 Parts

📊 Top Locations (Fr / Rr / Lh / Rh breakdown)

📊 By Model

📊 By Variant (Y/C/X/Q/D/R)

📊 Top 25 Parts Pareto

📊 Top 15 Parts × Detection Family

📊 Top 15 Parts × Location (where on the part)

🔥 Top 10 Parts × Location Heatmap

📊 Category Distribution

📊 Category × Family (where each category gets caught)

📊 Category × Location

📊 Category × Model

📊 Defects by Location

📊 Location × Family (where defects leak)

📊 Fr vs Rr (Front vs Rear) by Part

📊 Lh vs Rh (Left vs Right) by Part

Total defects
0
% of all defects
0%
Top part
—
Top location
—
Top family
—

📊 Selected slice — by Part

📊 Selected slice — by Location

📊 Selected slice — by Family

📊 Selected slice — by Model

📋 Matching Records (up to 50)

DateShiftModelPartLocationConcernFamilyResp AreaRecorded By

📊 Monthly Defects Trend

📊 Monthly by Detection Family

📊 Monthly by Top Parts

DateShiftModelVariantPartCategory LocationConcernFamilyResp AreaRecorded By

📋 Action Plan — based on cleaned data

🔴 Priority 0 — BUMPER (452 defects, 34.6%)

Bumper Fr + Rr alignment / fitment — 438 of these are self-tagged (TAGG)

Surfaces mostly in TAGG (438 production self-declared). Only 13 caught at QG. Distribution: Bumper-Rr = 229, Bumper-Fr = 221. Almost no QG catch suggests production catches them but they still leak.
Steps:
  • Pull all 452 Bumper records → cluster by variant (Y: 136, C: 108, std: 85, Q: 61, X: 33, D: 20, R: 9)
  • Investigate: gap/dimension mismatch? Mounting torque? Supplier batch?
  • Check bumper supplier dimensional reports for Y, C variants (largest volume)
  • 3-day side-by-side with T-P (RT) cell on bumper install
Impact: ~452 / 6 months · Effort: 2-3 weeks
🔴 Priority 0 — VIN STICKER (91 defects, 7%)

VIN Sticker missing or misplaced — 85 of these are TAGG (production self-declared)

Almost all VIN Sticker issues caught by production (TAGG = 85). This is a "forgot to apply" issue, not a "sticker quality" issue. Affects all variants: std (25), Y (23), C (23), Q (12), X (7).
Steps:
  • Audit current VIN sticker placement step in production sequence
  • Check if there's a verification step at end of each shift
  • Add VIN sticker to pokayoke check before vehicle moves to next station
  • Consider adding VIN sticker check to QG inspection
Impact: ~91 / 6 months · Effort: 1 week (simple SOP fix)
🟠 Priority 1 — HEADLINING (41 defects)

Headlining fitment — 40 of these are QG caught (in-process)

Distribution: Lh Fr = 23 (most common position). This means QG inspectors are catching it during inspection. Need to investigate why production isn't catching it themselves.
Steps:
  • Check headlining install jig + sequence
  • Verify headlining supplier dimensional reports
  • 3-day side-by-side between headlining station and QG
Impact: ~41 / 6 months · Effort: 1-2 weeks
🟠 Priority 1 — SIDE SKIRT (51 defects)

Side Skirt gap — distributed across HANE (24), TAGG (21), QG (6)

Side Skirt has the highest HANE leak rate of all parts. Distribution: Rh Fr = 21, Rh = 15, Lh Fr = 8, Lh = 7. Even though small volume, 47% leak to HANE (offline repair) is concerning.
Steps:
  • Same as previous Side Skirt recommendation: jig wear / supplier dim / torque sequence
  • Investigate why right-side gaps are more common (Rh Fr + Rh = 36, vs Lh Fr + Lh = 15)
Impact: ~51 / 6 months (but 24 leak to HANE) · Effort: 1-2 weeks
🟠 Priority 1 — PILLAR GARNISH (39 defects)

Pillar Garnish fitment — 39 of these are QG caught

Distribution: Lh Fr = 18, Rh Fr = 9, Lh Rr = 6. Almost entirely QG caught — like headlining, production not catching them. Likely install sequence or jig issue.
Steps:
  • Audit pillar garnish install SOP
  • Check if pillar garnish is installed before or after door assembly (sequence matters)
Impact: ~39 / 6 months · Effort: 1 week
🟡 Priority 2 — DOOR WEATHERSTRIP / BELT MOULDING

Door sealing parts — fitment issues, mostly QG caught

Door Weatherstrip (19) + Door Belt Moulding (17) = 36 defects, all QG-caught. Suggests install sequence or alignment issue at door assembly station.
Steps:
  • Review door sub-assembly cell SOP
  • Check weatherstrip/moulding supplier dimensional consistency
Impact: ~36 / 6 months · Effort: 1-2 weeks
🟡 Priority 2 — DATA QUALITY

92 records (_OTHER) + 297 records (Unspecified location) = 30% data noise

92 records have part=_OTHER, 297 records have location=Unspecified. This is the data quality issue. Can't trend or root-cause these without fixing the form.
Steps:
  • Update ATS Inspect form: Part + Location mandatory dropdowns
  • Block "_OTHER" selection without reason field
  • Re-train 3 active recorders (Fitry Abdillah, Mohd Tarmizi, Lokman Musa)
Impact: Better trend data · Effort: 1 week