Your Engineering AI Toolkit

From SOPs to AI Prompts

Every well-run engineering department has Standard Operating Procedures — step-by-step instructions for common tasks like failure analysis, incoming inspection, energy audits, and maintenance scheduling. These SOPs ensure consistency and capture institutional knowledge. But they are static documents. They do not analyse your specific data, they do not adapt to your context, and they cannot answer follow-up questions.

AI prompts are the next evolution of SOPs. A well-crafted prompt is like an SOP that comes alive — it takes your specific data, applies engineering logic, and gives you a tailored analysis or recommendation. And just like SOPs, good prompts should be documented, version-controlled, and shared across the team.

This chapter gives you a library of battle-tested engineering prompts. These are not generic AI tricks. They are domain-specific tools designed for the daily work of maintenance engineers, quality managers, design engineers, and plant managers in Indian manufacturing.

Building a Prompt Library

Why Structure Matters

A random prompt gives random results. A structured prompt gives consistent, useful output every time. The structure:

Prompt 1: Failure Analysis

You are a root cause analysis expert for industrial equipment in Indian manufacturing plants.

A [equipment type] at our plant in [location] has failed. Here are the details:
- Equipment: [make/model/rating]
- Age: [years in service]
- Last maintenance: [date and what was done]
- Failure symptoms: [what happened]
- Operating conditions before failure: [load, temperature, any unusual events]
- Historical issues: [previous failures or repairs]

Perform a structured root cause analysis:
1. List 5 possible root causes ranked by probability
2. For each, explain what evidence would confirm or rule it out
3. Recommend immediate actions (next 24 hours)
4. Recommend preventive measures to avoid recurrence
5. Estimate if this failure was predictable with condition monitoring

Consider Indian manufacturing context: power supply quality, ambient conditions, maintenance practices, spare parts availability.

Prompt 2: Anomaly Detection Interpretation

You are a condition monitoring specialist experienced with industrial machinery in Indian factories.

My AI monitoring system has flagged the following anomaly:
- Machine: [type, rating, running speed]
- Parameter: [vibration/temperature/current/pressure]
- Normal baseline: [value and pattern]
- Current reading: [value]
- Trend: [how it has changed over what period]
- Other parameters: [any correlated changes]

Interpret this anomaly:
1. What fault conditions could cause this specific pattern?
2. How urgent is this — can I run until next planned shutdown or should I stop now?
3. What additional measurements should I take to confirm the diagnosis?
4. If I must continue running, what monitoring frequency do you recommend?
5. What spare parts should I pre-order as a precaution?

Prompt 3: Quality Investigation

You are a quality engineering consultant specializing in machined components for Indian automotive and industrial markets.

We are experiencing increased rejections. Here is the data:
- Part: [description, material, critical dimensions]
- Normal rejection rate: [%]
- Current rejection rate: [%]
- Defect type: [dimensional/surface/material]
- When it started: [date or event]
- What changed recently: [new material batch, tool change, operator change, machine maintenance, weather change]
- Process parameters: [speeds, feeds, coolant, temperature]

Analyse this quality issue:
1. Most likely root causes (top 3) based on the defect type and timing
2. What data should I collect to confirm each hypothesis?
3. Immediate containment actions to stop defective parts from shipping
4. Short-term fixes (implement this week)
5. Long-term prevention (process or system changes)
6. BIS/ISO documentation requirements for this type of quality event

Prompt 4: Maintenance Scheduling Optimization

You are a maintenance planning expert for Indian manufacturing plants running 24/7 operations.

I need to plan maintenance for the next quarter. Here is my equipment list:
[Paste: Machine name | Last maintenance date | Condition status | Criticality (A/B/C) | Typical maintenance duration]

Constraints:
- Maximum 2 machines offline simultaneously
- Production demand: [high/medium/low] periods in [months]
- Festival shutdowns: [dates]
- Spare parts lead time: [weeks] from [supplier locations]
- Maintenance crew: [number] people per shift

Create an optimized maintenance schedule that:
1. Prioritizes by condition + criticality (not just calendar)
2. Groups related machines to minimize total downtime
3. Aligns major overhauls with planned shutdowns or low-demand periods
4. Ensures spare parts are ordered with sufficient lead time
5. Balances crew workload across the quarter

Prompt 5: Energy Audit Analysis

You are a BEE-certified energy auditor familiar with Indian industrial energy consumption patterns and tariff structures.

Here is our monthly energy data for the past 12 months:
[Paste: Month | kWh consumed | Maximum demand (kVA) | Power factor | Production output (units) | Energy cost (Rs)]

Also:
- Tariff structure: [time-of-day rates, demand charges, PF penalty/incentive]
- Major energy consumers: [list with rated kW]
- Operating hours: [shift pattern]
- Recent changes: [new equipment, production changes]

Analyse our energy performance:
1. Specific energy consumption (kWh per unit of output) — trend and benchmarking
2. Power factor analysis — are we paying penalties? How much can we save?
3. Maximum demand management — can we reduce peak demand charges?
4. Time-of-use optimization — which loads can shift to off-peak?
5. Top 3 energy-saving opportunities with estimated savings (Rs/year) and implementation cost
6. BEE PAT scheme compliance status if applicable

Prompt 6: Specification Review

You are a senior design engineer reviewing technical specifications for Indian manufacturing compliance.

Review this specification/drawing for:
[Paste specification details or describe the design]

Check against:
1. BIS standards applicability — which IS standards apply to this product/component?
2. Manufacturability — can this be made with standard Indian MSME capabilities (conventional CNC, standard heat treatment, locally available materials)?
3. Tolerance analysis — are the tolerances achievable and necessary? Flag over-specified dimensions.
4. Material availability — is the specified material readily available from Indian suppliers?
5. Testing requirements — what testing is needed for BIS certification or customer acceptance?
6. Cost implications — any specification choices that significantly increase cost without proportional performance benefit?

Provide output as: [Item | Issue | Recommendation | Impact (cost/time/quality)]

Open data/prompt-library-engineering.json in the code panel for an extended collection of 20+ engineering prompts organized by function — maintenance, quality, design, energy, safety, and compliance. Each prompt includes usage notes, example inputs, and expected output format.

Making Prompts Work in Your Team

Version Control Your Prompts

Just like engineering drawings have revision numbers, your AI prompts should too. When you find a prompt that works well:

Train Your Team in 30 Minutes

Most engineers are sceptical of AI until they see it solve their specific problem. Run a 30-minute session:

BIS/BEE Compliance Checks with AI

AI is particularly powerful for compliance checking because standards are structured, rule-based documents. A well-prompted AI can:

Open data/specification-analysis.json to see an example AI analysis of a mechanical component specification — showing how AI identifies over-specified tolerances, suggests material alternatives, and flags BIS testing requirements. Use this as a template for your own specification reviews.

Key Takeaways