Due Date

 

Status

Stakeholdersmarie.goavec@syensqo.combrian.bian@syensqo.com
Outcome


Contributors


Responsible


This page aims to translate the functional requirements into architecture tangible elements creating a engineering value perspective on the initiative - to assess business capabilities relevance versus architecture complexity.


Architecture Requirements Assessment

Quality AttributeRequirement - Architecture ConcernsArchitectural ComplexityBusiness Criticality/RelevanceBusiness Requirement Item
InteroperabilityHow "AI system" gather info from public & internal sources for scientists real-time evaluation?

Understanding Experimental Requirements and Information Collection
Usability

How "AI system" filter and displays complex data structure and diagrams?

Molecular Modeling and Simulation
UsabilityHow "AI system" supports/makes proposals to scientists on molecular modeling and simulation?

Molecular Modeling and Simulation
UsabilityHow "AI System" monitors digital reactor and formulation workstation processes to ensure accuracy/consistency of sample preparation?

Molecular Modeling and Simulation
ConsistencyHow "AI System" ensures accuracy/consistency of sample preparation?

Execution of Experimental Plan
UsabilityHow "scientists" monitor progress of digital reactor and formulation process in real time?

Execution of Experimental Plan
InteroperabilityHow "Central Control System" capture instrument data and promote it to Data Analysis System?

Sample Analysis and Data Processing
InteroperabilityHow "Central Control System" onboard/integrate new instruments? 

Sample Analysis and Data Processing
ConsistencyHow can "AI System" make recommendations for the experimental scheme to the scientists based on historical data and current results?

Experimental Optimization and Iteration
InteroperabilityWhat types of sensors "AI System" will connect and how it will monitor Environment and instrument parameters in real-time to schedule maintenance (to reduce downtime)?

Intelligent Management and Maintenance
SafetyHow "AI System" automatically checks existing group safety regulations/procedures and evaluates the safety of new instrument/processes/Management of Change to recommend additional safety monitoring parameters?

Safety and Compliance
UsabilityHow "AI System" allows scientists to access experiments database for evaluation?

Knowledge Management and Collaborative Work
UsabilityHow "Virtual Assistant" provides suggestions on best practices to scientists?

Safety and Compliance
Usability

How "AI System" enhances team collaboration, allows sharing data and offers personalized collaboration tips (recommendations)?

Knowledge Management and Collaborative Work
Usability

How "Reporting - Visualization System" keeps track on the regular work-flow to create analysis?

Automatic Report Generation and Environmental Control
Usability

How "AI System" automatically adjusts parameters based on experimental needs for optimal conditions?

Intelligent Management and Maintenance
Usability

How "Reporting - Visualization System" allows multi-project management?

Visualization System and Resource Scheduling
Interoperability

How "Voice User Interface (VUI)" allows personas in the lab to interact with Instruments and manage experiments?

Voice User Interface

Reference:

Architectural Concerns


    1. The "Big Data Hub" should be compliant with CyberSec OT/IT constraints.
    2. The solution must provide extensible and resilient interfaces mechanism for integrating with Syensqo LIMS, ELN and AI systems.
    3. Foundational or Pre-trained AI/LLM models consumed by Edge or Regular computing must be validated from the Syensqo AI standpoint and CyberSec constraints (+AI Risk).  
    4. Local network capacity must be adequately dimensioned to support the throughput of the number of sensors and their potentially complex and large data types capturing.
    1. Interoperability between systems (Machine Learning, Modeling-Simulation and LIMS/ELN) becomes critical to achieve efficiently the target seamless workflow. 
    2. Data integration should be taken into consideration to ensure the lineage and data consistency across different systems where users will perform their activities.
    3. Computer power is also critical for fine-tunning, training and real-time suggestions (here also low-latency network is a sensitive aspect to take into account).

→ It is also a concern to consider that eventually the solution used for Shanghai may not be generalizable or reusable in other regions, outside of China, due to aspects related to:

  1. Contract and legal for having the same vendors
  2. Interoperability with Syensqo Application Platform for experiments (LIMS, ELN)
  3. Data classification and data exchanges between regions

Utility Tree

The value engineering work on the user requirements allows to create such mind-map diagram so to visually capture the architectural significant requirement and its business criticality and complexity to be implementend.

Personas - Profiles

Environments

Components - Building Blocks   

High Level Design Architecture

There are three different views to describe the laboratory environment and the application/infrastructure architecture to accomplish the business needs.  








References

Functional Requirements

Smart Solution