How to Implement a Just-in-Time Inventory System for LED Therapy Manufacturing

We had $180,000 in inventory sitting in our warehouse. That was $180,000 of cash we couldn’t use for marketing, product development, or hiring. Worse, 15% of that inventory was for products we were about to discontinue — money sitting on shelves, slowly becoming obsolete.

We switched to a just-in-time (JIT) inventory system. Within 6 months, our inventory value dropped to $65,000 and our cash flow improved by $115,000. But JIT in LED therapy manufacturing comes with specific challenges that textbook JIT doesn’t cover.

—

## Why JIT Is Hard for LED Therapy

**Standard JIT assumptions:**
– Suppliers deliver frequently (daily or weekly)
– Lead times are short (1-5 days)
– Demand is predictable
– Quality is 100% — no incoming inspection needed

**LED therapy reality:**
– Suppliers are in China (minimum 30-day lead time)
– Components ship by sea (25-40 days transit)
– Demand is seasonal (Q4 is 40% of annual volume)
– Quality varies — incoming inspection is mandatory

**Our modified JIT approach:** We can’t achieve pure JIT with Chinese supply chains. Instead, we implement “JIT at the assembly level” — components arrive and move into assembly within 1-2 weeks, and finished goods ship within 3-5 days of production.

## The Demand Forecasting Foundation

JIT only works if you can predict demand. We use a three-method forecast:

**Method 1: Historical trend**
– Look at same-month sales from the previous year
– Adjust for year-over-year growth rate
– Weight: 40%

**Method 2: Pipeline analysis**
– Sum all confirmed B2B orders for the next 3 months
– Add projected DTC orders (based on current ad spend and conversion rates)
– Weight: 35%

**Method 3: Market signals**
– Amazon search volume trends (via Helium 10)
– Google Trends for LED therapy keywords
– Trade show timing (demand spikes after major trade shows)
– Weight: 25%

**Our 3-month rolling forecast accuracy:** ±18% (we over-forecast by 18% on average, which provides a safety buffer)

**Forecast review frequency:** Weekly (every Monday, the team reviews actual vs. forecast and adjusts)

## The Component Kanban System

We adapted the kanban system for components with long lead times:

**For each component, we maintain two inventory levels:**

1. **Order point:** The inventory level that triggers a new purchase order
2. **Safety stock:** The minimum inventory we’ll accept (the buffer against forecast errors and lead time variability)

**Example: LED chips (Epistar 633nm)**

– Average weekly consumption: 75,000 LEDs
– Factory lead time: 35 days (5 weeks)
– Safety stock: 2 weeks of consumption = 150,000 LEDs
– Order point: (5 weeks × 75,000) + 150,000 = 525,000 LEDs
– Order quantity: 750,000 LEDs (10 weeks of consumption)

**When LED inventory drops to 525,000, we order 750,000.** The order arrives in 5 weeks, by which time we’ve consumed 375,000 LEDs, leaving 150,000 in safety stock. The new order brings inventory back to 900,000, which provides 12 weeks of supply before the next order is needed.

**The kanban card (digital):** Our ERP system sends an alert when any component hits its order point. The procurement team reviews and places the PO within 24 hours.

**Components we manage on kanban:**
– LED chips (3 wavelengths)
– Silicone panels (2 sizes)
– Batteries (2 capacities)
– Flex PCBs (2 designs)
– USB-C ports
– Packaging materials

**Components we buy in bulk (not kanban):**
– Screws, adhesive, labels (too cheap to justify frequent ordering, minimum order covers 6+ months)
– MCU and driver ICs (long lead times, prone to shortages — we buy 6-month supply when available)

## The Assembly Schedule

**Our production schedule runs on a 2-week rolling horizon:**

**Week 1:** Produce based on confirmed orders + forecast for Week 2-3
**Week 2:** Produce based on confirmed orders + updated forecast for Week 3-4

**Daily production target:** 80-120 units/day (depending on demand)
**Production flexibility:** We can scale from 50 to 200 units/day with 2 days’ notice by adjusting overtime and temporary workers

**The scheduling rule:** Never produce more than 3 weeks of forecasted demand. If we produce ahead of forecast, we’re building inventory that ties up cash.

## Finished Goods Inventory Strategy

**We maintain two types of finished goods:**

1. **Make-to-stock (MTS):** Standard products kept in inventory for immediate shipment
2. **Make-to-order (MTO):** Custom or low-volume products produced only when an order is received

**MTS products and their stock levels:**

| Product | Daily Sales | Safety Stock | Target Inventory |
|———|———–|————-|—————–|
| GlowMask Pro | 25 | 175 (7 days) | 350 |
| GlowMask Lite | 15 | 105 (7 days) | 210 |
| GlowPanel Mini | 8 | 56 (7 days) | 112 |
| GlowCap | 5 | 35 (7 days) | 70 |

**We keep 7 days of safety stock for DTC products** (enough to cover demand during a production cycle). For B2B products, we keep zero finished goods inventory — we produce to order with a 35-day lead time.

## The Seasonal Buffer

**Q4 breaks JIT.** From October to December, demand is 2-3x normal. We can’t JIT our way through Q4 because:

1. Factory lead times increase (everyone’s ordering for Q4)
2. Sea freight capacity is limited and transit times are longer
3. Stock-outs during Q4 are catastrophic (lost revenue can’t be recovered)

**Our seasonal preparation:**

**August:** Place Q4 production orders with factory. Build 6 weeks of inventory for standard products.

**September:** Continue building inventory. Total stock: 8 weeks of forecasted Q4 demand.

**October:** Q4 demand begins. Ship from inventory while factory produces replacement stock.

**November-December:** Peak demand. Inventory drawn down. Factory produces at maximum capacity. Replenishment orders ship by air (more expensive but faster) if needed.

**January:** Post-Q4. Inventory is low. Normal JIT resumes.

**The seasonal inventory cost:** Approximately $45,000 in incremental inventory holding costs (financing + warehouse space + insurance). This is offset by avoiding stock-outs that would cost an estimated $120,000+ in lost revenue.

## Managing Component Shortages

JIT is fragile. A single component shortage stops production.

**Our shortage mitigation strategies:**

**1. Dual sourcing for critical components**
– LED chips: Epistar (primary) + NationStar (secondary)
– Batteries: Supplier A (primary) + Supplier B (secondary)
– USB-C ports: Supplier C (primary) + Supplier D (secondary)

**Cost of dual sourcing:** Approximately 5-8% premium on secondary-source components (lower volume = higher unit price). This is cheap insurance against a production stoppage.

**2. Buffer stock for long-lead-time components**
– MCU and driver ICs: 4-month supply on hand
– Custom flex PCBs: 6-week supply on hand
– These components have 12-16 week lead times and are prone to allocation during industry shortages

**3. Component substitution plan**
– For each critical component, we’ve qualified an alternative that can be substituted with minimal design change
– Example: If our primary MCU (STM32) is unavailable, we can switch to an alternative (GD32) with only a firmware recompile

**4. Advance ordering during allocation periods**
– When we hear about potential shortages (through industry contacts, news, or supplier warnings), we order ahead
– This builds buffer inventory at the cost of cash flow — but a production stoppage costs more than inventory carrying costs

## The Cash Flow Impact

**Before JIT:**
– Average inventory value: $180,000
– Inventory turns per year: 4.2
– Cash tied up in inventory: $180,000

**After JIT:**
– Average inventory value: $65,000
– Inventory turns per year: 12.8
– Cash freed up: $115,000

**What we did with the freed cash:**
– $50,000: Increased Q4 advertising spend (generated $180,000 in incremental revenue)
– $30,000: Hired a product development contractor (accelerated new product launch by 2 months)
– $35,000: Built cash reserve for component pre-ordering during shortage periods

**The cash flow benefit of JIT is often more valuable than the inventory cost savings.** Having $115,000 of working capital available instead of sitting on shelves gives you strategic flexibility.

JIT for LED therapy manufacturing isn’t textbook — it’s a modified approach that accounts for long lead times, seasonal demand, and Chinese supply chain realities. The goal isn’t zero inventory (that’s impossible with our supply chain). The goal is minimum inventory that supports reliable delivery without tying up excessive cash. Start with demand forecasting, implement kanban for your top 20 components, and adjust buffer levels based on actual experience. The cash you free up will pay for the effort many times over.