Master Deck Houston Summit 2026.pdf

/live_board/v2/downloads?uid=639d24c1-fdfe-4f05-8657-5649eb49c03d&redirect=true

Seeq Houston Summit March 3rd

Exclusive Summit Sponsor

Agenda

Time Session

10:30am – 11:00am Check-in & badge pick up

11:00am – 12:15pm Seeq Crash Course to AI, Kyle Clark

12:15pm – 12:35pm Customer Presentation: Indorama Ventures, Pablo Castillo

12:35pm – 12:55pm Break

12:55pm – 1:15pm Customer Presentation: Marathon Petroleum, Tim Sandford

1:15pm – 1:30pm BKO

1:30pm - 2:30pm Seeq Product Presentation, John Garramone

2:30pm – 4:00pm Happy Hour

Exclusive Roadshow Sponsor

AI Crash Course Prep!

Left side of the room

Network: Station 3 Password: 1919houston

Right side of the room

Network: Station3guest Password: 1919houston

Seeq Crash Course with AI

Kyle Clark, Senior Analytics Engineer

Crash Course Agenda

• Seeq Overview • AI Workshop in Seeq

Seeq Overview

Unique Problems Require Unique Expertise

What is the purpose of technology??

Capture

Distribute

Augment

Human Expertise © Seeq Corporation9

Data AI

Technology Alone Can’t Solve Million Dollar Problems

Your Experts Are The Key To Million Dollar Outcomes

Data AIExperts Query In Place

Data On Demand

Expert Enriched Training Data

Operationalize ML Predictions

Insight >> Decisions >> Outcomes

Purpose Built For Operational Analytics

Expert Knowledge Store

GenAI Accelerators

Your Experts

The Future Of Operational Intelligence Human-Centric AI Amplified Enterprise Ready

Seeq Operational Intelligence Platform

You are the Expert

The Goal

• See how you can solve tough industrial problems in Seeq

• …while getting a basic orientation to “what Seeq is and does”

• …with a focus on how AI accelerates value

Welcome to Seeq!

Sign into preview.seeq.dev • Username: Your Email Address • Password: Your Email Address all lower case

Navigate to your folder: My Folder > AI Assistant Training > Seeq Crash Course

(2 volunteers - what are you hoping to get out of this session?)© Seeq Corporation16

Who are you?

• You: Managing Engineer • Line: A4 • Product: Zyphorene (a high value

chemical)

What are you Doing?

1. Quickly identifying pump issues in the Zyphorene production line

2. Quantifying the total production each day • In the past this has been calculated infrequently, by hand,

using the weight of the transport truck measured at the gate.

Line A4

Flowserve Pump moves product

Meter Measures amount (flow)

Storage Tank

What Seeq Does, What Seeq Is

Connect, not copy Align Data Auto-updating

All Data in One Place

VANTAGE

VANTAGE© Seeq Corporation20

What Seeq Does, What Seeq Is

Knowledge Capture Reports & Dashboards Interactive Auto-updating

Connect, not copy Align Data Auto-updating

All Data in One Place

Self-Service Analytics No-code Tools + Formula Visualizations Notifications

Use Case &

Document VANTAGE

VANTAGE

VANTAGE© Seeq Corporation21

What Seeq Does, What Seeq Is

Knowledge Capture Reports & Dashboards Interactive Auto-updating

Connect, not copy Align Data Auto-updating

All Data in One Place

Self-Service Analytics No-code Tools + Formula Visualizations Notifications

Use Case &

Document VANTAGE

VANTAGE

AI for • Vantage Insights

• Analytics • Scripting • Learning/Skill-up

• Documentation

There are no rules!

• Let’s work together! • Feel free to comment, ask questions!

What we did

Knowledge Capture Reports & Dashboards Interactive Auto-updating

Connect, not copy Align Data Auto-updating

All Data in One Place

Self-Service Analytics No-code Tools + Formula Visualizations Notifications

Use Case &

Document VANTAGE

VANTAGE

AI for • Vantage Insights

• Analytics • Scripting • Learning/Skill-up

• Documentation

What’s Next

Knowledge Capture Reports & Dashboards Interactive Auto-updating

Connect, not copy Align Data Auto-updating

All Data in One Place

Self-Service Analytics No-code Tools + Formula Visualizations Notifications

Script & Code Simplify data sci Customize add-ons Big ML integration

Scale Scale Analytics over similar assets

Use Case &

Document VANTAGE

VANTAGE

AI for • Vantage Insights

• Analytics • Scripting • Learning/Skill-up

• Documentation

What’s Next

Knowledge Capture Reports & Dashboards Interactive Auto-updating

Connect, not copy Align Data Auto-updating

All Data in One Place

Self-Service Analytics No-code Tools + Formula Visualizations Notifications

Script & Code Simplify data sci Customize add-ons Big ML integration

Scale

“Work the Problem”

Scale Analytics over similar assets

Enterprise Insights Monitor, triage, respond to business-level issues

Use Case &

Document

VANTAGE

AI for • Vantage Insights

• Analytics • Scripting • Learning/Skill-up

• Documentation

Indorama Marina Donaldson and Pablo Castillo

Discover Indorama Ventures

Marina Donaldson Director of Sustainability,

Global EHSQ Leadership Team

Pablo Castillo Mechanical Engineer

Indovinya, Dayton Site

Boiler Deaerator Optimization Sustainability Project

Company Snapshot

Indorama Ventures is a global chemical company based in Thailand.

Indorama Ventures is a global chemical company

based in Thailand. Founded by Group CEO Aloke

Lohia over 30 years ago, Indorama Ventures

expanded quickly as a family-owned business to

become a leading global PET manufacturer.

In 2010, Indorama Ventures listed on the

Thailand Stock Exchange and reset its vision

to become a global chemical company by

building a global portfolio of integrated assets

across its chosen petrochemical value chain.

Today, it has four business segments: Combined

PET (CPET), Indovida (Packaging), Fibers,

and Indovinya. ~1.5 billion

EBITDA (FY 2024)*

~11 billion

(capital employed)*

Presence in

32 countries *update as of Apr 2025

Consolidated revenues of

$15.4B (FY 2024)

~18 million tons

(capacity)*

114 Manufacturing locations *update as of Apr 2025

About

25.000 employees

*Data as of financial year 2024

Our customers help us grow as a market leader

Ethoxylation company globally

#1 in PET staple fiber ASEAN

Lifestyle

#2 in BiCo

fiber

Hygiene

#2 in airbag

yarn

Mobility

#2 in tire cord

fabric

Mobility

Fibers

#1 Non-ionic surfactants Producer in Americas

#1 Americas’ home care ingredients providerIndovinya

#1 PET producer globally #1 Recycled PET

producer globally

The only shale-integrated player in the West

#1 Crop solutions provider in Americas

Indovida

CPET

Packaging leadership in emerging markets#1 Market

position in ASEAN #2 Market

position in Africa

Indovinya chemistry touches our lives every day

Paints

Ink

Auto oil lubricants

Biocides

(sanitizers)

DetergentsToothpastes

Surface cleaners

Shampoos

Nutrition and health

Herbicides

White Goods

Indovinya: Global Footprint

15 manufacturing

locations

7 R&D

centers

Presence in

10 countries

~ 3000 employees

India

Australia

China

APAC

R&D & Tech Centers

Denotes Sales Presence

Industrial Units

USA

Brazil

Americas

Mexico

Uruguay

Belgium

EMEA

Indorama Ventures | Sustainability "Sustainability in Indovinya is about doing the right thing, The right thing for our people, our communities and our customers."

Mr. Alastair Port Executive President at Indovinya

INDOVINYA 1st Place Sustainability Category

Pablo Castillo

Mechanical Engineer

Indovinya, Dayton Site

Boiler Deaerator Optimization Sustainability Project

INDORAMA VENTURES - INDOVINYA

Originally constructed in 1975 for the production of agricultural surfactants, the Dayton plant is a specialty batch facility that produces a broad range of surfactants, EO/PO derivatives, and blends while providing scale-up capabilities for new Indovinya products.

Dayton, Texas

4Improve

Analysis Consolidation

and automatic reporting

to support decision-

making and continuous

monitoring.

3Measure /

Analyze

Analysis and

development of metrics,

enable optimization and

identify of improvement

opportunities.

21 Define

Collection and

organization of essential

information for analysis,

ensuring reliable and

structured data.

Control

Implement long term,

automation & Projects

to resolve this fully

without relying on

constant reports.

ANALYTICAL PROCESS

PROCESS BACKGROUND & OVERVIEW

System is designed to recover >80% of condensate for boiler reuse. As wastewater is transported offsite, condensate losses create immediate financial and environmental impact.

Maximizing recycle drives measurable value:

• Fuel Optimization - Reduced energy demand for feedwater heating

• Water Stewardship - Lower fresh makeup consumption

• Cost Control - Reduced wastewater hauling volume and disposal cost

• Emissions Reduction - Smaller operational carbon footprint

Wastewater

Air

BFW

Condensate losses (Bypassing due to Deaerator limitation)

Makeup Water

Condensate Return

HP steam

Process

Heating

Heat losses

B - Boiler Primary

Steam losses

Deaerator

A - Boiler

Waste Water Generation Components = Rain + Cooling Tower System + Process + Condensate Losses

Background: Condensate Recovery — Strategic Utility Lever

Problem Statement: Advanced analytics and soft-sensor modeling reveal condensate pinch points — turning hidden energy losses into actionable insight that reclaims heat and strengthens sustainability performance.

SCHEMATIC – Steam & Condensate System

Improve ControlMeasure / AnalyzeDefine

Challenges Monitoring KPI Seeq Utilization

System Performance Visibility Steam, Condensate and Feedwater correlated with Energy, Water use, and Wastewater Generation Integrated dashboards to identify performance drifts

High Energy & Water Use Excursion events correlated with Gas intensity, makeup and wastewater flow Identify events and quantify losses

Measuring Performance Gas use, Makeup water, condensate recovery %, DA stability Establish baselines, Rolling KPI’s and KPI Scorecards

Operational Variability Startup / transition events with DA / Condensate behavior Identify high variability periods and Quantify Impact

KEY CHALLENGES IN STEAM & CONDENSATE SYSTEM PERFORMANCE

Improve ControlMeasure / AnalyzeDefine

(2) Flow?

Energy

(2) Manual Valve /No Instrumentation

Heat losses

PROCESS BACKGROUND & OVERVIEW | Closer Look Dayton Steam & Condensate System

SCHEMATIC – Steam & Condensate System

Process

Heating Steam losses

BFW

Condensate losses (Bypassing due to Deaerator limitation)

Makeup Water

Condensate Return

Deaerator

Wastewater

Air

A B

Maximizing Condensate Re-Use

A. Boiler Firing Rate (Real-Time Energy Input) • Converted fuel gas flow into normalized firing rate (%). • Applied rated boiler heat input to calculate real-time energy consumption

and operating load.

B. Make-Up Water Estimation • Converted deaerator tank level into gallons using tank geometry. • Used level rate-of-change to calculate true makeup water entering the boiler

system. • Enabled monthly tracking of makeup demand and condensate recovery

performance. C. Energy & Sustainability View

• Integrated engineered signals into a unified energy balance and recovery view.

• Identifies recovery gaps, losses, and improvement opportunities across the system.

Unmeasured energy, therefore unmanaged cost. Turning hidden energy, into actionable insights with Seeq.

Improve ControlMeasure / AnalyzeDefine

ANALYTICAL & DIAGNOSTIC APPROACH

• Identify all existing signals / tags

• Identify gaps from instrumentation, validate data quality, timestamps, and units

Data Mapping

• Convert units and normalize tags

• Define conditions

• Create “virtual” signals for missing sensors

Data Engineering

• Select reference window between boiler outages (2025)

• Exclude abnormal/ outage periods to set baselines

Identify Baselines

Improve ControlMeasure / AnalyzeDefine

Signal Identification & Data Mapping | A. Boiler Firing Rate ANALYTICAL & DIAGNOSTIC APPROACH

Improve ControlMeasure / AnalyzeDefine

• Review and add existing signals

• Identify gaps due to limited instrumentation

• Validate data quality and continuity

• Utilize formula tool to convert units and create new calculated

signals.

• Document signals and calculations supporting sustainability KPIs

Boiler Firing Rate and Heat Input

Baseline

• SCFM = ACFM ( 𝑃𝑎

𝑃𝑠𝑡𝑑 ) × (𝑇𝑠𝑡𝑑

𝑇𝑎 )

• Heat Input Rate ( 𝐵𝑇𝑈

ℎ𝑟 ): Flow Rate (𝑓𝑡3

ℎ𝑟 ) × Calorific Value ( 𝐵𝑇𝑈

𝑓𝑡3 )

• Firing Rate (%) = ሶ𝑄𝑎𝑐𝑡

ሶ𝑄𝑟𝑎𝑡𝑒𝑑 × 100

Equations

Monthly Heat Input MMBtu

ANALYTICAL & DIAGNOSTIC APPROACH

Improve ControlMeasure / AnalyzeDefine

Step 1. Immediate Stabilization - Operational Controls (Short-Term Mitigation)

• Implement procedural adjustments to manage system pressure

• Maintain active monitoring and precise control until structural repairs are finalized

• Estimated feedwater temperature based on pressure–temperature correlation

• Translated deaerator tank level into gallons through tank geometry modeling

• Calculated actual makeup water entering the boiler using the rate of change in level

• Adjust manual bypass valves to manage depressurization events

Condensate System MATBAL & Modeling

Signal Identification & Data Mapping | B. Make-Up Water Estimation

Estimated Make up water (when feed water valve opens)

Step 2. Engineering Controls — System Improvements (Long-Term Mitigation) Automate bypass pressure control, remove condensate return bottlenecks, and redesign manual operations currently required— enabling higher, stable condensate recovery within system limits.

Overall Energy Balance & Sustainability KPIs

ANALYTICAL & DIAGNOSTIC APPROACH

Improve ControlMeasure / AnalyzeDefine

• Energy View

• Integrated engineered signals into a unified energy balance and recovery view

• Identifies recovery gaps, losses, and improvement opportunities across the steam system

Signal Identification & Data Mapping | C. Energy & Sustainability View

CO2 Generation

Sustainability View When the additional condensate is returned: • Reduced makeup water demand • Lower fuel required for feedwater heating • Reduced wastewater hauling • Lower emissions footprint • Direct operating cost reduction

• Enthalpy Ratio = ℎ𝑔−ℎ𝑔𝑤,ℎ𝑜𝑡

ℎ𝑔−ℎ𝑓𝑤,𝑐𝑜𝑙𝑑

• ሶ𝑚𝑓𝑢𝑒𝑙 × 𝐻𝐻𝑉 × 𝜂 = ሶ𝑚𝑠𝑡𝑒𝑎𝑚 ℎ𝑠𝑡𝑒𝑎𝑚,𝑜𝑢𝑡 − ℎ𝑓𝑤,𝑖𝑛

• 𝐶𝑂2 lb = Natural Gas (MMBtu) × 116.9*1 *1 Standard Emission Factor (U.S. EPA)

Equations

Condensate carries the highest thermal value in the system. Alternatively, makeup water enters cooler and requires significant energy input.

Analyze | Insights & Value

ANALYTICAL & DIAGNOSTIC APPROACH

Improve ControlMeasure / AnalyzeDefine

Baseline for “expected performance” for deeper analysis

• Define key operational states

• Maintenance periods, abnormal operation and steady state conditions

• 1 year period: 2024 boiler outage to 2025 boiler outage selected as the primary baseline

• Data validation for consistency in gas flow behavior, makeup water usage, and DA operation

• Establish a baseline to serve as the reference point for performance improvements and sustainability impacts.

Baseline Period

Waste Water Generation Components = Rain + Cooling Tower System + Process + Boiler Makeup

Other Water (CT, Process, A-Boiler, etc)

61% B-Boiler Makeup

26%

Unit Washes 13%

Utilities & Process Waste Water Generation 2025

Baseline

ANALYTICAL & DIAGNOSTIC APPROACH

Improve ControlMeasure / AnalyzeDefine

• Natural Gas • 11.3% gas losses estimated based on underheated water

enthalpy ratio • Avg monthly gas usage equating to over 400 MMBTU/month gas

losses • Estimated annual emissions of 254.7 tCO2 e/yr

• Makeup Water & Wastewater • Avg Makeup water over 100K gallons/month • Water is obtained from well so there is no cost associated with

usage. • Estimated annual emissions of 31.6 tCO2 e/yr from contract

wastewater trucks

Insights & Sustainability Value

Estimated savings if feedwater temp is increased

Monthly Makeup Water Totalized

Sustainability Opportunities

ANALYTICAL & DIAGNOSTIC APPROACH

Improve ControlMeasure / AnalyzeDefine

• Operational updates implemented to improve condensate return, with early signs

of higher feedwater temperature and partial gas and water savings expected

• Ongoing KPI tracking in Seeq to verify performance against the baseline

• Capital proposal for a surge/flash tank to further improve condensate

management

• Regular cross-functional reviews with operations to share insights, discuss

excursions, and align actions

Summary

CONCLUSION

• Seeq allowed us to turn raw process data into clear insights by creating soft signals and

applying calculations that exposed issues that were previously unmeasured and therefore

unmanaged

• Establishing baselines and KPIs gives us a clear way to verify improvements and sustain

performance over time

• Early results show meaningful operational gains, and ongoing monitoring will confirm the

full energy, water, and sustainability benefits

• This work supports better decision making, strengthens reliability, and advances the

sustainability goals of our steam and condensate system

48© Indorama Ventures

This presentation and its content (“Material”) is proprietary to Indorama Ventures Public Company Limited (“Indorama Ventures”) and/or its affiliates (collectively, the “Group”) and may not be, in whole or in part, reproduced or disclosed, published, distributed or released to any other person or to the public domain unless the prior written consent from the Group is obtained. In addition, this Material may only be used for the purpose expressly stated herein by Indorama Ventures and may not be used for any other purposes.

No representation or warranty or undertaking, express or implied, is made by the Group as to the accuracy or completeness of the information set forth herein and neither Indorama Ventures nor the Group (or any representatives including, without limitation, its and their directors, shareholders, officers, employees, agents (“Representatives”) assume any responsibility whatsoever related hereto.

In addition, this Material may contain “forward-looking” statements of the Group that relate to future events including, without limitation the conditions and prospects of the specific industry and the macro economics as a whole which are, by their nature, subject to significant risks and uncertainties. All statements, including, without limitation, those regarding the future financial position and results of operations, strategy, plans, objectives, goals and targets, future developments in the markets where the Group participates or is seeking to participate and any statements preceded by, followed by or that include the words “target”, “believe”, “expect”, “aim”, “intend”, “will”, “may”, “anticipate”, “would”, “plan”, “could”, “should, “predict”, “project”, “estimate”, “foresee”, “forecast”, “seek” or similar words or expressions are forward-looking statements. Such forward-looking statements involve known and unknown risks, uncertainties and other important factors beyond the Group control that could cause the actual results, performance or achievements of the Group to be materially different from the future results, performance or achievements expressed or implied by such forward-looking statements. These forward-looking statements are based on numerous assumptions regarding the Group present and future business strategies and the environment in which the Group will operate in the future and are not a guarantee of future performance.

Such forward-looking statements speak only as at the date of this presentation, and neither Indorama Ventures nor the Group assume any duty or obligation to supplement, amend, update or revise any such statements. In addition, neither Indorama Ventures nor the Group hereby make any representation, warranty or prediction that the results anticipated by such forward-looking statements will be achieved.

As such, no information contained herein may be relied upon as a promise or presentation as to the past, present or future of Indorama Ventures or the Group and use of this Material therefore is subject to informed assessment and independent evaluation of the person to which this Material is disclosed. Further, the receipt of this Material shall not be taken to constitute the giving of investment advice by any of Indorama Ventures or the Group (and/or their respective Representatives) nor render the recipient a client of any such persons for the purpose of any applicable rules or regulations governing investment business or otherwise.

This Material does not constitute an offer to sell or the solicitation of an offer to buy securities, nor will there be any sale of securities in any state or jurisdiction in which such offer, solicitation or sale would be unlawful prior to registration or qualification under the securities laws of any such jurisdiction. No offering of securities will be made except by means of a prospectus meeting the requirements of the applicable securities laws, or an exemption therefrom.

Disclaimer

Questions ?

THANK YOU!

PABLO CASTILLO

3rd March 2026

Presented By:

Break See you back at 12:55pm!

Meet the Attendees!

Marathon Petroleum Tim Sandford, Senior Refinery Engineer

AI in Energy 2026

Case Study: Predictive Maintenance: Anomaly Detection for Smarter Refinery Operations at Marathon

Tim Sandford – Refining Operations Research

Agenda

Background – MPC Enterprise Overview

Asset Health Monitoring (AHM)

Process Transients

Enterprise Unit Monitoring with DRA

Fault Tree Analysis

Q&A

55© Seeq Corporation

Background – MPC Integrated System

Refining

⚫ Approximately 3 million barrels per

calendar day of refining capacity from 15

refineries in 12 states

⚫ Each refinery contains thousands of

pieces of equipment with varying levels of

instrumentation for maintenance and

optimization

⚫ Technical Services, Maintenance,

Operations, and other teams monitor

hundreds of data points that they deem

important

⚫ Individual contributors only have the

bandwidth to actively monitor a handful of

these data points each day/week/month – Can lead to process deviations being missed,

causing an upset or equipment damage that was

wholly avoidable

56© Seeq Corporation

Asset Health Monitoring (AHM)

⚫ Asset Health Monitoring (AHM): strategic

enterprise-wide initiative designed to

improve asset reliability, process safety,

and operational performance across MPC

– Gather data from refining assets

– Deliver predictive and diagnostic insights

⚫ Goal: let the data work for us by

highlighting deviations from normal

operation

AHM – Process Transients

⚫ One of the key capabilities enabled by

AHM is the concept of a Process Transient

⚫ Process Transient (PT): short-duration,

non-steady-state operating behaviors that

indicate a transition between operating

regimes

– Can be intentional or unintentional

– Fast-acting, high fidelity process data

⚫ Goal: detect, review, and take preventive

action on PTs before they cause a bigger

problem

DRA and Enterprise-Wide Unit Monitoring

⚫ MPC combines the analytics engine powering DRA’s anomaly detection with the

concept of PTs to identify meaningful shifts in behavior from normal operation

⚫ PTs are added to DRA, then are organized by refinery, unit, and technology as-

needed. Example include:

– Coker heater tubeskin temperatures

– FCC riser/reactor temperature

– Alkylation unit reactor outlet temperatures

⚫ Leverage the power of DRA’s autonomous analytics engine to serve meaningful

insights to refining leadership and individual users/teams and prevent upsets or

failures before they occur

DRA and Enterprise-Wide Unit Monitoring

Fault Trees

Dynamic Fault Trees

⚫ Link directly to live process data

⚫ Enable rapid identification of likely root causes

⚫ Support faster, more effective responses to process issues

Autonomous Analytics Integration

⚫ Leverages DRA’s autonomous analytics and operating-fitness

⚫ Model system signals in parallel & series

⚫ Provides proactive, automated reporting & notifications

Anomaly Detection Models

⚫ DRA offers 3 types of models for anomaly detection:

⚫ M1: high sensitivity model that ingests short, medium, and long-term history of

process data

– Vast majority of tags analyzed at MPC use this model

⚫ M2: medium sensitivity model that ingests medium and long-term history of

process data

– Better suited for noisier/more variable data

⚫ M3: low sensitivity model that ingests long-term history only of process data

– Best model for extremely noisy or dynamic/variable data

The Seeq Solution

Operationalizing autonomous analytics at enterprise scale

Bridging Detection to Decision

• Converts DRA anomaly signals and process transients into contextualized engineering insight

• Accelerates root-cause identification across interconnected process variables

• Shortens the cycle from deviation detection to corrective action

Enterprise Surveillance Without Engineering Bottlenecks

• Scales seamlessly across thousands of tags, units, and entire refinery networks

• Eliminates per-asset model development and ongoing retuning

• Enables exception-based monitoring without increasing staffing requirements

Process Transient Intelligence

• Automatically detects and quantifies short-duration, non-steady-state events

• Standardizes transient review across sites, technologies, and operating teams

• Transforms transient events into governed, repeatable workflows

Dynamic Fault Tree Enablement

• Integrates live time-series data directly with fault-tree logic

• Validates the most probable root causes in real time

• Supports faster, more confident operational decision-making

Enterprise Impact

• Strengthens asset reliability and enhances process safety

• Reduces avoidable upsets, equipment stress, and unplanned downtime

• Improves leadership visibility and operational consistency across the refining portfolio

Keys to Success

⚫ Asset instrumentation at Scale

– Wireless Microsensors

⚫ Data Quality and Fidelity

⚫ Ingestion of Data

⚫ Results are Executable

– Board Operators, Reliability Engineer

Historians • Autonomous Machine Learning

• Installs in a day, no engineering maintenance required

• Scales quickly across thousands of tags, assets, and process areas

• Capabilities available on both cloud and on-premise versions

• Archive system for all process data

• Useful tool to visualize process diagrams, tags

• Both on-prem, cloud capabilities

• Custom first-principal models

• Model creation is resource intensive, and models must be

adjusted / re-calibrated periodically

• Cloud access typically required

Digital Twins

Advanced Process Control (APC) • Addresses specific process performance

• Limited to controlled variables only, < 10-20% of total variables

• Long resource intensive process, requiring process control experts

Time Series AI (DRA)

• Requires “normal” & “failure” signatures for each asset

• Model creation is resource intensive, and models require tuning

& re-calibration periodically

• Each asset needs to be modeled individually

Asset Performance Management (APM)

• Advanced statistical toolbox

• Engineer dependent: Analysis is to be done manually by users

• Often there is steep learning curve to obtain useful insights

• Cognitive bias

Self-Service Analytics

Digital Landscape

Questions?

Thank you!

BKO Presentation Ziad Katrib and Dr. Bin Liu

Common Model Connected-Data and AI-Agent-Ready Platform for Everything

Ziad Katrib Co-founder ziad@bkoai.com

Dr. Bin Liu AI Technology Manager bin@bkoai.com

Common Model (CM)

• Insight Layer (Top) • Knowledge-enriched context • AI-driven deep analysis

• Refined Layer (Middle)

• Agentic contextualization • CM Canvas • CM Knowledge Graph Explorer

• Raw Layer (Foundation)

• CM Broker for CRUD operations • Connected data from diverse sources: Seeq, SQL,

Neo4j, Databricks, AVEVA, Digital Twins, P&IDs, etc • AI-driven P&ID digitalization

Common Model (CM)

• Cross-System Data Connectivity

• Agentic Context Processing

• Engineering Context & P&ID Integration

• Validation-to-Production Governance

Common Model Canvas

• Intuitive SME Interface

• Unified Search & Analysis

• Visual Graph Connectivity

• Context-Driven Workflow

CM Knowledge Graph Explorer

• Complex Hierarchy Navigation

• End-to-End Traceability

• Impact Analysis

• Root Cause Analysis (RCA)

• AI-Agent Integration

Thanks! Welcome to BKO AI booth and discuss more! info@bkoai.com +281-221-2836 www.bkoai.com

Seeq Product Update

John Garramone, Product Manager

If I’m shining, everybody gonna shine.

Lizzo

Decision Intelligence Flywheel

ScaleAnalyze Operationalize

Condition Monitoring

TriageInvestigate Decide

Act

Quantify

Seeq's Product Strategy

Seeq Intelligence AI Workers AI Chief of Staff AI Profit Agents

Now Next Future

Seeq's Product Strategy

Seeq Intelligence

Seeq Enterprise Workflow at scale, Monitoring at scale

Governance at scale, Models at scale

AI Workers

Live in chat, or triggered by monitoring activity,

integrated into the AI ecosystem

AI Chief of Staff

Personal highly-capable Industrial AI Concierge

for every user

Curiosity at scale, Action at scale

AI Profit Agents

AI Agents autonomously identify cost and profit drivers, making proactive suggestions

Now Next Future

Seeq's Product Strategy

Seeq Intelligence

Seeq Enterprise

Seeq Analytics

Workflow at scale, Monitoring at scale

10x ROI through rapid amplification of analysis outcomes

and built-in workflows

Governance at scale, Models at scale

20x ROI through sustained, widespread deployment of

continuous improvement drivers

AI Workers

Live in chat, or triggered by monitoring activity,

integrated into the AI ecosystem

AI Chief of Staff

Personal highly-capable Industrial AI Concierge

for every user

Expanded Tooling

Curiosity at scale, Action at scale

30x ROI through broad adoption of powerful what-if modeling

and tailored optimizations

AI Profit Agents

AI Agents autonomously identify cost and profit drivers, making proactive suggestions

Expanded Possibilities Expanded Audience

Now Next Future

Seeq's Product Strategy

Seeq Intelligence

Seeq Enterprise

Seeq Analytics

Workflow at scale, Monitoring at scale

10x ROI through rapid amplification of analysis outcomes

and built-in workflows

Governance at scale, Models at scale

20x ROI through sustained, widespread deployment of

continuous improvement drivers

AI Workers

Live in chat, or triggered by monitoring activity,

integrated into the AI ecosystem

AI Chief of Staff

Personal highly-capable Industrial AI Concierge

for every user

Expanded Tooling

Empower SMEs to do more: Statistics, Multivariate, Batch

Curiosity at scale, Action at scale

30x ROI through broad adoption of powerful what-if modeling

and tailored optimizations

AI Profit Agents

AI Agents autonomously identify cost and profit drivers, making proactive suggestions

Expanded Possibilities

Empower SMEs to solve for more: Time-Series + Tabular Data, work

order integration

Expanded Audience

Empower more users to be SMEs: Zero to Hero in minutes

Now Next Future

Impact Reports

From Analysis to Outcome

• Capture clear stories with quantified impact

• Illustrate exactly how insight translates into value

• Enable broad visibility of use cases & outcomes achieved

Scale What Works

• Create an enterprise-wide catalog of solutions

• Unlock discovery and reuse of proven value plays

• Elevate best practices through discoverability

Impact You Can't Ignore

• Showcase success stories that align to key stakeholders

• Inform investment decisions and executive storytelling

• Prioritize resources and champion what’s working

Machine Learning Tools

PLS Predict quality and process outcomes

External ML Score custom ONNX Models

Prediction or Detection Integrates Databricks, Azure ML, Sagemaker

SOM Detect operating shifts

PCA Detect multivariate anomalies

Clustering Discover natural operating patterns

Isolation Forest Identify rare and unexpected behavior

Detection

Spot behavior shifts

Prediction

Forecast performance

External ML

Bring your own model

Production-ready ML that drives earlier insight and better operational decisions.

Machine Learning Tools

With Seeq, your team goes from reactive to proactive:

Statistics native in Seeq – no add-on, no exports Build, validate and operationalize statistics in one live environment. Deploy once, scale enterprise -wide

Today: You have to go looking for problems and risks Manual checks, disconnected add-ons, and gains that fade back into variability

Seeq tells you when to act Run-time SPC alerts surface deviations automatically – before they become exceptions.

Status: Early access, first feature set general availability late Q2

Proactive Monitoring | Native Statistics | Scale Across Operations

Statistical Process and Quality Control (SPC / SQC)

Statistic Summaries in Seeq

Better Together | Enterprise Accelerators

Accelerating data-driven decisions, at scale

Package Calculations

“I’ve built best practice calculations for one unit. I need them to be reusable

across all facilities.”

Models

Contextualize Data

“I have a list of tags. I need to organize them into a scalable,

discoverable, and manageable structure.”

Tree Builder

Scale your Analysis

“I need to deploy my calculations to 1000s of assets, processes,

or batches.”

Scaling Tables

Investigate, Analyze, Act

“I need to efficiently monitor my asset/process health to investigate, prioritize, and

make data-driven decisions”

Better Together | Enterprise Accelerators

Accelerating data-driven decisions, at scale

Package Calculations

“I’ve built best practice calculations for one unit. I need them to be reusable

across all facilities.”

Models

A fresh approach for subject-matter- experts to package analytics solutions for

reusability and scaling

Scalable, Sustainable Analytics Model

Inputs Outputs 𝑓(𝑥)

𝑓(𝑥)

𝑓(𝑥) 𝑓(𝑥)Key Differentiators Sharing best-practice analytics

Standardizing on scaled deployments Faster onboarding for all users

Models Demo Video

Better Together | Enterprise Accelerators

Accelerating data-driven decisions, at scale

Package Calculations

“I’ve built best practice calculations for one unit. I need them to be reusable

across all facilities.”

Models

Contextualize Data

“I have a list of tags. I need to organize them into a scalable,

discoverable, and manageable structure.”

Tree Builder

A powerful UI to create, manipulate and manage asset hierarchies for the use cases

you want, when you need it

Build-your-own hierarchy

Key Differentiators Easier data navigation and discovery

Accelerated scaling of analytics Crowd-sourced by your experts

Tree Builder Demo Video

Better Together | Enterprise Accelerators

Accelerating data-driven decisions, at scale

Package Calculations

“I’ve built best practice calculations for one unit. I need them to be reusable

across all facilities.”

Models

Contextualize Data

“I have a list of tags. I need to organize them into a scalable,

discoverable, and manageable structure.”

Tree Builder

Scale your Analysis

“I need to deploy my calculations to 1000s of assets, processes,

or batches.”

Scaling Tables

An innovative way for SMEs to flexibly scale Seeq calculations across 1000s of assets or more, for their use

case, when and how they need to

try else elseScale Across Assets

Key Differentiators

Rapid calculation scaling and iteration Managing 1000s of calculations

Handling asset asymmetry with fallback rules

Scaling Table Demo Video

Better Together | Enterprise Accelerators

Accelerating data-driven decisions, at scale

Package Calculations

“I’ve built best practice calculations for one unit. I need them to be reusable

across all facilities.”

Models

Contextualize Data

“I have a list of tags. I need to organize them into a scalable,

discoverable, and manageable structure.”

Tree Builder

Scale your Analysis

“I need to deploy my calculations to 1000s of assets, processes,

or batches.”

Scaling Tables

Investigate, Analyze, Act

“I need to efficiently monitor my asset/process health to investigate, prioritize, and

make data-driven decisions”

Vantage Demo

Better Together | Enterprise Accelerators

Accelerating data-driven decisions, at scale

Package Calculations

“I’ve built best practice calculations for one unit. I need them to be reusable

across all facilities.”

Models

Contextualize Data

“I have a list of tags. I need to organize them into a scalable,

discoverable, and manageable structure.”

Tree Builder

Scale your Analysis

“I need to deploy my calculations to 1000s of assets, processes,

or batches.”

Scaling Tables

Investigate, Analyze, Act

“I need to efficiently monitor my asset/process health to investigate, prioritize, and

make data-driven decisions”

What Powers Seeq Intelligence

Agent Extensibility

Bi-Directional

Agent to Agent / Custom Agents

Agent Q

AI Analyst

Synthesis of Information

Document Access

Procedures, Manuals, Investigations, etc.

Closing Shaista Mallik, Industry Principal Oil and Gas

Exclusive Summit Sponsor

Early bird registration closes March 16th!

Meet the Attendees!

Slide 1: Seeq Houston Summit Slide 2: Exclusive Summit Sponsor Slide 3: Agenda Slide 4: AI Crash Course Prep! Slide 5: Seeq Crash Course with AI Slide 6: Crash Course Agenda Slide 7: Seeq Overview Slide 8 Slide 9: What is the purpose of technology?? Slide 10: Technology Alone Can’t Solve Million Dollar Problems Slide 11: Your Experts Are The Key To Million Dollar Outcomes Slide 12 Slide 13: Seeq Operational Intelligence Platform Slide 14: You are the Expert Slide 15: The Goal Slide 16: Welcome to Seeq! Slide 17: Who are you? Slide 18: What are you Doing? Slide 19: Line A4 Slide 20: What Seeq Does, What Seeq Is Slide 21: What Seeq Does, What Seeq Is Slide 22: What Seeq Does, What Seeq Is Slide 23: There are no rules! Slide 24: What we did Slide 25: What’s Next Slide 26: What’s Next Slide 27: Indorama Slide 28 Slide 29 Slide 30 Slide 31 Slide 32 Slide 33: Indorama Ventures | Sustainability Slide 34 Slide 35: INDORAMA VENTURES - INDOVINYA Slide 36: ANALYTICAL PROCESS Slide 37: PROCESS BACKGROUND & OVERVIEW Slide 38: KEY CHALLENGES IN STEAM & CONDENSATE SYSTEM PERFORMANCE Slide 39: PROCESS BACKGROUND & OVERVIEW | Closer Look Slide 40: ANALYTICAL & DIAGNOSTIC APPROACH Slide 41: ANALYTICAL & DIAGNOSTIC APPROACH Slide 42: ANALYTICAL & DIAGNOSTIC APPROACH Slide 43: ANALYTICAL & DIAGNOSTIC APPROACH Slide 44: ANALYTICAL & DIAGNOSTIC APPROACH Slide 45: ANALYTICAL & DIAGNOSTIC APPROACH Slide 46: ANALYTICAL & DIAGNOSTIC APPROACH Slide 47: CONCLUSION Slide 48: Disclaimer Slide 49 Slide 50 Slide 51: Break Slide 52: Meet the Attendees! Slide 53: Marathon Petroleum Slide 54: AI in Energy 2026 Slide 55: Agenda Slide 56: Background – MPC Integrated System Slide 57: Asset Health Monitoring (AHM) Slide 58: AHM – Process Transients Slide 59: DRA and Enterprise-Wide Unit Monitoring Slide 60: DRA and Enterprise-Wide Unit Monitoring Slide 61: DRA and Enterprise-Wide Unit Monitoring Slide 62: Fault Trees Slide 63: Anomaly Detection Models Slide 64: The Seeq Solution Slide 65: Keys to Success Slide 66: Digital Landscape Slide 67 Slide 68 Slide 69: BKO Presentation Slide 70: Common Model Slide 71: Common Model (CM) Slide 72: Common Model (CM) Slide 73: Common Model Canvas Slide 74: CM Knowledge Graph Explorer Slide 75 Slide 76: Seeq Product Update Slide 77 Slide 78: Decision Intelligence Flywheel Slide 79 Slide 80 Slide 81 Slide 82: Seeq's Product Strategy Slide 83: Seeq's Product Strategy Slide 84: Seeq's Product Strategy Slide 85: Seeq's Product Strategy Slide 86: Impact Reports Slide 87 Slide 88: Machine Learning Tools Slide 89: Machine Learning Tools Slide 90 Slide 91: Statistical Process and Quality Control (SPC / SQC) Slide 92: Statistic Summaries in Seeq Slide 93 Slide 94 Slide 95: Better Together | Enterprise Accelerators Slide 96: Better Together | Enterprise Accelerators Slide 97: Models Slide 98: Models Demo Video Slide 99: Better Together | Enterprise Accelerators Slide 100: Tree Builder Slide 101: Tree Builder Demo Video Slide 102: Better Together | Enterprise Accelerators Slide 103: Scaling Tables Slide 104: Scaling Table Demo Video Slide 105: Better Together | Enterprise Accelerators Slide 106: Vantage Demo Slide 107: Better Together | Enterprise Accelerators Slide 108 Slide 109: What Powers Seeq Intelligence Slide 110 Slide 111 Slide 112 Slide 113: Closing Slide 114: Exclusive Summit Sponsor Slide 115: Early bird registration closes March 16th! Slide 116: Meet the Attendees!

Master Deck Houston Summit 2026.pdf

Item Type: pdf