5. Scoping the Project

This article contains the following sections:


By considering the customer’s project landscape and building a thorough project description, you can determine which of our product offerings is the best fit for the customer:

  • Visibility Offering: Enterprise Manufacturing Visibility (EMV)

  • Analytics Offering: Enterprise Manufacturing Analytics (EMA)

Visibility Offering: Enterprise Manufacturing Visibility (EMV)

The Enterprise Manufacturing Visibility (EMV) product:

  • Streams in-scope data to our manufacturing analytics platform for real-time visibility into factory and machine performance
  • Integrates with the customer’s existing technology systems
  • Can be incorporated with the customer’s additional digital projects

With EMV, the customer can use digital technology to drive operational improvement across its enterprise, laying the foundation for the use of manufacturing analytics at scale.

EMV Customer Value Matrix




Increase capacity utilization and output by identifying production issues and analyzing plant utilization across the supply chain in real time.

Monitor regulatory information, birth certificate, and track and trace by maintaining compressive and accurate information about products throughout the entire lifecycle.

Operationalize digital manufacturing initiatives with tools for creating custom applications to transform workflow.


Reduce scrap by analyzing process, part, and quality data to discover the root cause of complex plant problems.

Improve throughput by identifying bottlenecks and performance drivers impacting yield.

Drive continuous process improvement through deeper insights into quality/product issues and work-in-progress inventory levels to pinpoint the most impactful initiatives.


Increase first-pass yield by capturing real-time first-pass yield statistics and immediately identifying variances that affect quality.

Control processes by setting parameters to strict tolerances and establishing alerts when a process is out-of-bounds.

Predict asset failures through the use of historical data, trends, and manufacturing-specific algorithms to manage asset failures proactively.

EMV Applications and Features

The EMV product provides the following applications and features for the customer’s facilities, machine types, and machines:

  • Global Operations View: A real-time enterprise view of machine operations and output across plants
  • KPI Dashboard: Key plant and machine performance metrics by facility and machine type across the customer’s network of manufacturing facilities
  • Data Visualization: Visibility into real-time and historical machine and sensor data, providing an integrated view of the customer’s global operations
  • Visibility Applications: Real-time visibility of facility and machine performance: output, availability, and downtime
  • AI Data Pipeline: Easy-to-use configurator that creates data models of facilities and machines
  • Cloud Storage: Secure storage and retrieval of the customer’s machine data, optimized to work with any cloud provider
  • Translate and Transmit: Intelligent edge processing to secure, optimize, and normalize machine and factory data for streaming into the cloud

Analytics Offering: Enterprise Manufacturing Analytics (EMA)

The Enterprise Manufacturing Analytics (EMA) product delivers advanced analytics for improved productivity, quality, and supply chain optimization.

EMA Customer Value Matrix




Create a multi-plant scoreboard that consolidates performance metrics to enable a centralized view of asset availability and OEE.

Analyze real-time factory floor data to enable business model change and/or evaluate performance.

Access real-time supplier/contract manufacturer production data.


Access a real-time deep dive into production from anywhere on the factory floor.

Compare past production runs to optimize performance.

Analyze historical data prior to downtime to determine process improvement opportunities.


Detect, track, and diagnose line stops and starved/blocked conditions.

Monitor and remove process deviation.

Perform forensic analysis with contextualized historical data instead of taking the line down to reproduce problems.

EMV and EMA Product Comparison

Use the chart below to determine which product offers the most suitable applications and features for the customer’s facilities, machine types, and machines.

Determining Scoping Milestones and Decision Points

In any engagement, it is important to agree with the customer upon realistic milestones and decision points. Think of the milestones as the various deliverables from the statement of work, while the decision points are the dates tied to those milestones.

You will be responsible for holding onsite meetings with the customer aimed at helping them solve particular problems or issues using Sight Machine products. You should ask thorough questions of the customer to fully understand what their specific manufacturing concerns are. Your goals should be the following:

  • To define the scope of the project
  • To define how the team will measure the success of the project
  • To show how Sight Machine product features can help resolve the defined problems

During the stages of the implementation process, you will assess whether or not various items fit within the project scope as well as how much effort or customization they would require.

The stages you should consider include:

  • Data Intake: Make the assessment after evaluating data and building the customer value map.
  • Data Modeling: Risks may not be revealed until after data intake but, to every extent possible, evaluate what you see.

You should evaluate the risks and categorize situations as the following:

  • In-Product Fast Installation: These are items that fit well within the scope and offer few impediments to success. Overall, data is well-organized and easy to ingest. These items should comprise the first set of milestones, and generally will be in scope for an EMV/EMA project.
  • Fee-Based Data Preparation and Modeling Services: These are items that have high complexity and would need to be negotiated with the customer, usually as part of the professional services agreement. Beware that this is usually where projects become overly complicated and run aground.
  • Not Ready For Digital Transformation: These are items that we should probably not undertake as part of the project because of the complexity involved. Usually, these items involve a high level of customer and sometimes vendor involvement, both of which add time and cost to the project. All of these items should be discussed with the customer in the context of DRI as being out of scope for the project.

Technical Readiness Requirements Matrix


In-Product Fast Installation

Fee-Based Data Preparation and Modeling Services

Not Ready For Digital Transformation

Data Intake

Customer data must be accessible. For example:

· Data in an OPC UA-compatible “tag” format (from PLC, MES, or historian)

· Well schema-ed SQL database

· CSV (time series data or serial number with quality measurements)

· Consulting on data quality best practices

· Creation of specific plugins and modules

· Conversion of non-tag connectivity (Modbus, wire protocol, etc.)

· Standardized bespoke data formats

· Data is not immediately accessible (e.g., behind an airgap).

· Data is in undocumented, proprietary formats.

· Data has been manually entered without appropriate controls.

Data Modeling

· Data is reported for each source in a regular, real-time cadence, with clear definition of fields.

· Customer SME who understands the meaning of the data can help guide the modeling.

· Data cleaning: building relationships using timestamps or inconsistent serialization

· Reconciling data rhythm cadences

· Data formatting: organizing, ordering, labeling

· Data has no recognizable mapping to operational outcomes.

· There is an inability to match data (no serialization).

· Data has bad or inconsistent timestamps.

· There is no automated way of repeatedly capturing data.

The following table describes how each potential failure might affect the overall project.

Potential Failure

Effect of Failure

Critical data modeling component is unavailable (e.g., timestamp, serial, part info).

Unable to create contextualization.

No data is available from critical process areas (e.g., a machine is not sensored).

Unable to uncover appropriate causal inputs.

Data population is inconsistent (e.g., missing data points).

Unreliable analytics (sampling error).

There is poor data quality (e.g., unable to ingest directly without cleanup).

Inability to meet project deadlines.

There is low data volume (e.g., small population).

Poor prediction capability.

There is extremely high data volume.

Poor platform performance.

Data flow is interrupted.

Unreliable analytics (sampling error).

Manufacturing line or process data is misunderstood.

Unreliable data contextualization.