Seeds of Time: Technology Forecasting

Meet Your Facilitator - Prof. Bala Ramadurai

Meet Your Facilitator - Bala Ramadurai

Entrepreneur, Professor, Author and Innovation Coach

• Marie Curie Research Fellow at Politecnico di Milano, Milan, Italy
• PhD from Arizona State University, USA (Materials)
• BTech from IIT Madras, India (Metallurgy)

Prof. Bala Ramadurai - Professor

• IIT Madras, Chennai, India
• Universidad Panamericana, Mexico City, Mexico
• Symbiosis Institute of Business Management, Pune, India

Bala Ramadurai - Author

Bala Ramadurai - Innovation Coach

Bala Ramadurai - Entrepreneur

Spirelia

Innovate with Confidence

Srishti OS — The AI-Native Innovation Platform

Spirelia Innovation Pvt. Ltd.

"Innovate with Confidence, Powered by Insight"

S R I S H T I — The AI-Native Innovation Platform

🔭

Utkarsha

Technology Forecasting
& Roadmapping

MVP LIVE

🔬

Uttishta

AI Inventor's
Assistant

IN PIPELINE

💡

Karmaja

Structured Innovation
& Idea Generation

VISION

🛡️

Narasimha

IP Capture &
Protection

VISION

⚙️

Ananta

Rapid
Prototyping

VISION

Courses Taught

Design Thinking

Human Centred Design of products and services

Innovate Like a Boss

Systematic Innovation using methodologies like TRIZ, Brainstorming

Technology Forecasting

Quick Sensing of the Room

• Show of hands — Who here is primarily in R&D? Production? Testing? Management?
• Show of hands — Who has been involved in forecasting or roadmapping before?
• Shout it out — AI is not an acceptable answer: what technology will quietly dominate defence by 2035?

If you can look into the seeds of time, And say which grain will grow and which will not

• From Macbeth, Act I, Scene 3

Why Tech Forecasting Now?

1975: DRDO forecast India's missile needs
MTCR denial came too late — foundations were built
"No technology control regime can derail our missiles"

When we forecasted — we withstood denial.

Image: Wikimedia Commons (CC BY), DRDO/Govt. of India

Intro to Technology Forecasting

What is Technology Forecasting?

Technology

• Hardware (The physical)
• Software (The virtual)
• Orgware (The rules/knowledge)

Technology = Hardware + Software + Orgware

Forecasting

• time
• scale

Time

Scale

Why Technology Forecasting?

Technologies operate in a context

• Organizational context

Why Technology Forecasting?

• Move from invention to innovation
• Increase effectiveness of our efforts
• Work with limited resources

Technology Forecasting is Strategic Decision Making

How do we forecast technologies?

• Step 1: Formulate the tech forecasting project
• Step 2: Model the system
• Step 3: Act on the data
• Step 4: Transfer the results

Note - Model here is curve fitting of parameters of the forecast, in the context of the tech forecasting project

Step 1: Formulate the project

• Pose the questions that you seek answers to in the organizational context
• The questions should contain these three:
  • what
  • when
  • where
• Get the buy-in of the beneficiaries/decision makers
• Example - Will vacuum forming technologies be needed in the future (20 years, 2013-2033) for Whirlpool products (Refrigerators) in factories in Europe?

Step 2: Model the system

• Understand how the system works

• Determine the critical parameters

  

Step 3: Act on the data

• Use critical parameter data from databases and model them into an s-curve (logistic)

  

Step 4: Transfer the results

• Present the answers of questions from Step 1 to the beneficiaries/decision makers

• Example - Will vacuum forming technologies be needed in the future (20 years, 2013-2033) for Whirlpool products (Refrigerators) in factories in Europe?

• Answer:
  • Yes, vacuum forming will be needed for Whirlpool in factories in Europe

Remember FOR.M.A.T

• FORmulate
• Model
• Act
• Transfer

Note - Model here is curve fitting of parameters of the forecast in the context of the tech forecasting project

Source - G. Cascini, B. Ramadurai, et al. https://handbook.format-project.eu

An example - Weapon Locating Radars

Note - This example for illustrative purposes only, not to be taken literally

Sample Case: Future-Ready Weapon Locating Radars

Organizational Context

• Design & develop advanced radar systems
• Support Tri-Services & other defense forces
• Establish indigenous production capabilities

Why It Matters

• Informs R&D investments
• Enhances force protection
• Enables effective counter-battery operations

Technology Forecasting

FORmulate

• Identify the system - WLR
• Formulate the questions to be answered in the organisation context
  • Example - How accurately and quickly will future weapon locating radars be able to detect and respond to hostile artillery across diverse terrains and contested electronic environments over the next 10-15 years (2026-2041)?
• Tip - You could break down a complex questions into sub-questions
• Identify the roles of the team members - One leader per stage

Scenarios

Pick a project from your own domain that you want to perform a forecast on

Teams

• Pair or team up with someone in a similar domain as yours

Tasks for You

1. Get to know your team
2. Get to know your theme
3. Perform FORmulate steps

Time remaining:

Self-evaluate: FORmulate

• [ ] 1 mark for clearly identifying your system
• [ ] 1 mark for defining the organizational context
• [ ] 2 marks for a well-formed forecast question containing what, when, and where
• [ ] 1.5 marks for breaking the question into sub-questions
• [ ] 1.5 mark for identifying team roles (one leader per stage)

Model

Note - Model here is curve fitting of parameters of the forecast in the context of the tech forecasting project

Water Case study

The hydro-distribution infrastructure, conceptualized to effectuate conveyance of sanitized aqueous sustenance to anthropological end-consumers, encompasses interdependent subsystems. The hydraulic propulsion mechanism, energized through insufficiently robust electrochemical current provisioning, receives tributaries from suboptimally adequate hydrological repositories. Said apparatus channels aqueous material to a contaminant-elimination subsystem—this interrelationship being satisfactorily operational—maintained through ameliorative interventions by a servicing mechanism team, itself reliant upon insufficient electrochemical energization. A pecuniary transaction apparatus interfaces with end-consumers for informational dissemination and resource dispensation, both linkages presently inadequate. The corporate entity exercises proprietorial dominion over purification, maintains the servicing team, remunerates a commercial promotion division, while treated aqueous output is hydrostatically conveyed to end-consumers through an insufficiently performant pipeline.

Quiz

1. What are the weak points in the system?
2. What does the company control?
3. What are some external factors that are not easily controlled?

Visual report

Quiz

1. What are the weak points in the system?
2. What does the company control?
3. What are some external factors that are not easily controlled?

Imagine conveying your best idea(s) through words alone

Words can be ambiguous. We need a visual tool to capture our system.

Function Analysis

I speak to you

Pen writes on paper

Visually

How do we describe a system visually?

Step 1: List Objects of the System

Step 2: Connect the Objects

Step 3: Add verbs to the relationships

Step 4: Classify the relationships and add a legend

Step 5: Write the function

Source - Kamarudin et al, Procedia Engineering 131 ( 2015 ) 1064 – 1072

Source - https://triz-journal.com/case-study-with-triz-allocation-system-for-a-processing-machine/

Functional Analysis - The procedure

1. List Objects of the System
2. Connect the Objects
3. Add verbs to the relationships
4. Classify the relationships and add a legend
5. Write the function

For our case: WLR

WLR Objects — Internal Subsystems

WLR Connections (Illustrative)

WLR - FA diagram

Task - Functional Analysis

1. Start with your current system
2. Perform functional analysis
3. What (all) would you like to improve in the system? - Optional for this workshop

Time remaining:

Self-evaluate: Functional Analysis

• [ ] 1 mark for listing all key objects of your system
• [ ] 1 mark for connecting the objects with relationships
• [ ] 1 mark for adding verbs (actions) to each relationship
• [ ] 2 marks for classifying relationships (useful, harmful, insufficient)
• [ ] 1 mark for writing at least one complete function statement
• [ ] 1 mark for identifying what you would like to improve

Recap

This is not part of our workshop, but is the next step

• What all would you like to improve in the system?
• Answer: top 2 parameters
  • Target Localisation Error (metres)
  • Time-to-Engagement (seconds)

Act

4 seasons model

Quantitative curve-fitting of critical parameters

Note - This example for illustrative purposes only, not to be taken literally

Why do we normalize?

• The S-curve models cumulative growth (improvement over time)
• But our raw data shows error decreasing from 150m to 15m
• We invert: convert "error reducing from 150 to 15" into "improvement growing from 0 to 135"
• Formula: Improvement = Initial Value - Current Value
• This gives us a monotonically increasing series that the logistic curve can fit

Normalization before we fit logistic

| Year | Target Localisation Error Margin (m) |
|------+--------------------------------------|
| 1995 |                                    0 |
| 2000 |                                   30 |
| 2005 |                                   60 |
| 2010 |                                   90 |
| 2015 |                                  115 |
| 2020 |                                  130 |
| 2023 |                                  135 |  

Fit the data into Spirelia's s-curve software - UtkarshaSigmoid

• Copy and paste the data into UtkarshaSigmoid
• Note down the predictions, tm = time of maximum growth rate
• Note down the value of K (estimated saturation point or minimum achievable error)

https://tf.balaramadurai.net/s-curve.html

Median and forecast

• K (Saturation Point / Minimum Error): Approximately 5 metres
• tm (Time of Maximum Growth Rate): Approximately 2007

Strategic Interpretation: So What?

• K = 5m means the technology is approaching its physical limit for this generation
• tm = 2007 means we passed the inflection point nearly two decades ago – we are deep in the diminishing-returns zone
• Strategic implication: Diminishing returns on current WLR approach
• Decision: Invest in next-generation approaches (e.g., multi-sensor fusion, AI-assisted tracking) rather than incremental improvement of existing radar

Tasks for you

1. Identify the variable for s-curve fitting from your system
2. Use the data table provided by the facilitator (or use the WLR example data)
3. Normalize the data (Improvement = Initial Value - Current Value)
4. Draw a grid on plain paper, plot the normalised data, and sketch the S-curve
5. Mark tm (steepest point) and K (saturation level) on your graph
6. Discuss with your team: What does your tm and K tell you strategically?

Time remaining:

Self-evaluate: Act

• [ ] 1 mark for identifying the right variable for s-curve fitting
• [ ] 1 mark for correctly normalizing the data
• [ ] 1 mark for drawing a grid and plotting the data points
• [ ] 1 mark for sketching a smooth S-curve and marking tm and K
• [ ] 2 marks for interpreting what tm and K mean for your system's future

Transfer

Checklist for Transfer

• ✅ Answer the Question to be Forecasted
• ✅ Executive summary
• ✅ Report
• ✅ Presentation

Executive Summary Format

1. Strategic Question: What we set out to forecast
2. Key Finding: The S-curve parameters (tm, K) and what they mean
3. Implication: What this means for our R&D roadmap
4. Recommendation: Specific actions to take
5. Confidence & Limitations: Data quality and assumptions

Some tips for Transfer

• Lead with the strategic question, not the math
• Refresh your beneficiaries' memory about the case
• Use a maximum of 5+/-2 objects in every slide
• Try to convey your message visually

Handling Pushback

• "What if the forecast is wrong?"
  • All forecasts have uncertainty – the value is in structured thinking, not precision
  • The S-curve provides a baseline; scenario planning handles the rest
• "Our technology is different"
  • Every technology follows lifecycle patterns – the question is where you are on the curve
• "We don't have enough data"
  • Even 4-5 data points can reveal a trend; the alternative is pure guesswork

Tasks for you

1. Complete the presentation
2. Invite your beneficiary from the other team
3. Present the results to the beneficiary
4. Seek feedback
5. Improve presentation

Final Self-Evaluation: Your FORMAT Journey

• 16-20: You've internalized FORMAT - ready to forecast!
• 10-15: Solid foundation - one more practice cycle will sharpen your skills
• 0-9: Review the concepts and try the activities again with your team

AI in Technology Forecasting

How can AI assist in Technology Forecasting?

• FORmulate: AI can help scan literature, identify trends, and refine research questions
• Model: AI can extract system components and suggest functional relationships
• Act: AI can gather time-series data, fit curves, and interpret parameters
• Transfer: AI can draft executive summaries, generate reports, and prepare presentations

AI is a Co-pilot, Not a Replacement

• Domain expertise remains essential – AI doesn't know your organizational context
• AI accelerates the process, but you must validate the outputs
• The FORMAT methodology provides the structure; AI provides the speed

Demo: Utkarsha – FORMAT Forecaster

• AI-powered assistant that guides you through the entire FORMAT workflow
• Conversational interface – ask questions, get structured guidance
• Built-in tools: STEEP Analysis, Scenario Builder, Impact Assessment, Strategic Options
• Generates professional reports (PDF, PPTX, DOCX) from your forecasting session

Live Demo

Let's walk through a forecasting project together using Utkarsha

Key Utkarsha Features

• AI Coaching: Contextual suggestions at every FORMAT stage
• Document Analysis: Upload papers, reports – AI extracts relevant data
• ATRA Roadmapping: AI-suggested technology architecture and dependency mapping
• S-Curve Fitting: Integrated UtkarshaSigmoid for quantitative analysis
• Works online (cloud AI) or offline (local AI for sensitive environments)

Conclusion

We learnt this

Your Commitment

• What is one system you will apply FORMAT to within the next 30 days?
• Write it down now – and share with your team
• The best way to learn forecasting is to do a forecast

Now What?

①

Download the FORMAT handbook

http://handbook.format-project.eu

②

Register for SWAYAM/NPTEL Course

Registration will be open soon!

Your SWAYAM Instructors

Prof. Dmitry Kucharavy	Prof. Bala Ramadurai

Acknowledgments

- Emacs - this old editor rocks! - https://www.gnu.org/software/emacs/
- Spacemacs - this new configuration is space age - https://spacemacs.org
- Org Mode - I run my life using this - https://orgmode.org
- Reveal.js - cool presentation script - https://revealjs.com/
- Org-reveal package - lets me live in org-mode - https://gitlab.com/oer/org-re-reveal
- Org-teaching - original codebase for this presentation - https://gitlab.com/olberger/org-teaching
- Plantuml - for all the cartoon work - https://plantuml.com/
- Hugo - for converting into static html - https://gohugo.io
- Gitlab - for hosting my website and the presentations - https://gitlab.com

Acknowledgments

• Conversations with
  • Prof. Dmitry Kucharavy, EM Strasbourg Business School, University of Strasbourg, France
  • Prof. Gaetano Cascini, Politecnico di Milano, Milan, Italy
  • Dr. Murali Loganathan, Spirelia
• European Union Marie Curie People Project

Meet Your Facilitator - Prof. Bala Ramadurai

Love to keep in touch

https://balaramadurai.net/about/

bala@balaramadurai.net

http://in.linkedin.com/in/balaramadurai

My LinkedIn

https://www.linkedin.com/in/balaramadurai/

The Future Forecasters from the Past

2025, ADE, Bengaluru

2025, DRDO scientists at IIT Bombay

2026, AVML, Avadi, Chennai

Oh and one more thing

Selfie

2026-DRDO-TF