Sanjib Das: Failures Aren’t Static: Make Spare Decisions Dynamic (Part 5)

(Part 5 of our series with reliability engineer Sanjib Das)

In Part 1, Reliability Expert Sanjib Das showed why spare parts decisions should be mathematical, not emotional.
In Part 2, we examined why “just in case” inventory quietly destroys reliability and cash.
In Part 3, we explored how the right spares shorten downtime — they don’t prevent failure, but they compress the consequence window.
In Part 4, we showed how multiple sites can cut cost and risk by pooling medium-critical spares globally.

Now, in the final part of this series, we look ahead.

Because everything you’ve read so far leads to one realisation:

A static spreadsheet cannot keep up with a dynamic plant.

And Sanjib says this explicitly.

The problem with spreadsheet-based spare parts logic

Sanjib built his first tool in Excel, and it worked.
But even he reached its limits:

• it relies on manual updates
• engineers forget to refresh assumptions
• lead times change
• failure behaviour changes
• assets are modified
• work orders add new evidence
• criticality must be revisited
• supplier performance evolves

And most importantly:

“This is not a static thing… failure is dynamic.”

He is clear that spare parts decisions need to move with the plant.

Why static models drift out of reality

Three things change continuously in an operating environment:

1. Failure behaviour changes over time

Sanjib gives a simple example:

“This seal fails… but we upgraded it and it didn’t fail for five years.”

After five years without a failure, a part that once deserved a “high” score may become “medium”. That changes stocking rules entirely. But Excel won’t catch this unless someone manually updates the sheet.

2. Lead time assumptions decay quickly

Real suppliers rarely match yesterday’s data.
Sanjib emphasises that lead time is a key factor in the scoring model, but it’s also one of the most volatile:

• new suppliers appear
• old suppliers consolidate
• global logistics shift
• customs times fluctuate
• local sourcing becomes possible

Static spreadsheets freeze these assumptions.

3. Work orders create continuous new evidence

Every new failure, inspection and root cause analysis adds to the real reliability picture.

Sanjib stresses:

“If you have data, use the field data.”

But using field data means feeding it back into the model, something spreadsheets rarely enforce.

What dynamic models enable

A dynamic model does what Excel cannot: it updates itself.

Sanjib describes exactly what he started to do:

“I got someone to write the Python code… it will pull the data, it will analyse this, it will change.”

This is his blueprint for the future:

• pull data from SAP, Maximo or other ERPs and CMMS
• integrate updated failure history
• refresh asset criticality
• update lead times
• recalculate stocking levels
• alert engineers when thresholds shift

In other words: Dynamic stocking = the model changes when the world changes.

What AI brings — and what it doesn’t

Sanjib is enthusiastic about AI, but realistic:

“AI and machine learning makes our life easier… because we have limitation for our human capacity.”

He is not advocating replacing planners. He is advocating keeping planners informed.

AI helps by:

• identifying pattern shifts
• flagging unusual failure frequencies
• detecting deviations in supplier delivery
• surfacing parts with changing risk profiles
• recalculating stocking based on live data

And because it doesn’t forget, get distracted or deprioritise, it maintains focus:

“Sometimes because of the firefighting… we lose focus. AI keeps constant focus.”

This is the heart of his argument.

Where SPARROW.Stock fits in

A dynamic planning model only works if the underlying inventory data stays accurate and this is where SPARROW.Stock completes the picture.
Most plants struggle with catalogue inconsistencies, missing attributes and slow-moving records that never get updated. Over time, this erodes trust in the data and forces teams back into spreadsheets.

SPARROW.Stock provides a modern, intuitive interface for maintaining spare parts catalogues and inventory levels, making accuracy part of the daily workflow rather than an annual clean-up exercise. Engineers can quickly validate quantities, correct discrepancies, flag obsolete items and record preservation requirements. The result is a continuously reliable view of what is in stock, what is moving, and what should be phased out: a foundation that allows SPARROW.Plan to make confident, dynamic recommendations over time.

Everything Sanjib attempted manually, from Excel formulas to Python scripts, sits at the foundation of SPARROW’s vision: Spare parts decisions that are risk-based, transparent and continuously updated, not trapped in a static spreadsheet.

The competitive imperative

Sanjib’s closing message in the interview is blunt:

“If you do not use [AI], and others use it… you cannot compete.”

He illustrates it with a cost example:

• one plant produces something at EUR 10
• another, using AI-driven optimisation, produces it at EUR 8

This is not about technology hype. It is about operational economics.

Plants that adopt dynamic stocking will:

• hold less inventory
• free up working capital
• reduce preservation workload
• recover faster from failures
• operate with fewer surprises
• make reliability more predictable

Plants that don’t will keep firefighting. If you want to quantify the impact for your own sites, our ROI Calculator gives a quick estimate of the working-capital savings, preservation cost reduction and downtime avoided through optimised stocking and pooling.

Where this journey leads

Across these five articles, Sanjib has walked us through the full arc of modern spare parts thinking:

1. Stop Stocking on Gut Feel: Make it mathematical.
2. The Hidden Cost of “Just in Case”: Inventory isn’t free.
3. Planned vs Unplanned: Good stocking shortens downtime.
4. One Global Spare vs Three Local: Pooling cuts cost without adding risk.
5. Excel Is Static. Failures Aren’t: Dynamic models keep pace with reality.

For reliability and maintenance leaders, the takeaway is clear:

Static thinking creates static stock. Dynamic plants need dynamic stocking.

And the organisations that embrace this shift will run more reliably, more safely and more competitively than their spreadsheet-bound peers.

If you want to see how SPARROW works in practice, book a demo with our team. We also provide spare parts data quality assessments based on a sample of your own data, a fast master-data audit that reveals duplicates, inconsistencies and harmonisation gaps across your catalogues. It’s the simplest way to understand your current data issues and the value of fixing them.