Integrated Systems 
Science

Thrust lead:        Marianthi Ierapetritou, Rutgers
# of faculty:         12
Disciplines:        Chemistry, Chemical Engineering,

                              Computer- aided Engineering

THRUST D

DEM (EDEM, DEM Solutions) simulations of particles in a rotating drum granulator at various times, colored by liquid fraction.

press to zoom

Model predictive control of drug concentration

press to zoom

Application of DEM and gPROMS based modeling software

press to zoom

DEM (EDEM, DEM Solutions) simulations of particles in a rotating drum granulator at various times, colored by liquid fraction.

press to zoom
1/4
Current Thrust D projects
D1
Sensing
D2
 Integration
D4
Real-time Process Management
D5
Integrated Design & Optimization
Summary

The overall challenges in integrated systems science deal with the identification of critical material and/or process parameters that should be monitored for control, sensing the “right” information we need in real-time, developing robust reduced-order models that use real-time data to provide decision support for design, control, optimization, and exceptional events management. Additional components include an informatics framework to keep track of all the data-information-knowledge in the system, and an integrated system of hardware and software that can interface with the process and execute all this efficiently. The specific key challenges are: (i) the lack of analytical methods with sufficient resolution, speed, and reliability for measuring micro- and nano-structure properties in a “noisy” environments, (ii) lack of a systematic framework to store, access, manipulate, manage and use enormous amounts of complex and varied data, information, and models for optimal, real-time decision-making in design, control, and operations, (iii) lack of an integrated system of hardware and software elements for efficient and robust real-time execution of model-based algorithms for control and operations of continuous processes, (iv) lack of understanding of process behavior and its response to operating conditions and design parameter changes.  


Thrust D projects balance the need to advance fundamental issues in manufacturing systems science with the need to develop the technological applications of these methods in the various test-beds. The model development has focused on required unit operations such as feeding, mixing, blending, roller compaction, milling, tableting etc. The tools developed in this thrust cut across the different scales of product development and manufacturing, incorporating the structure-composition-properties models developed in Thrusts A-C, and across the three test-beds.

Goals

The central goal of this thrust is to design, scale-up, monitor, control and optimize structured organic particulate manufacturing systems and processes using predictive models and advanced computational methodologies in an integrated framework to guide optimal decision making. The main objectives are:

  1. Implement advanced science-based sensor technologies for effective process monitoring and state estimation in the manufacture of engineered composite products

  2. Optimize the performance of manufacturing processes by developing and integrating advanced model-based informatics, supervisory control strategies, design and scale-up, and real-time optimization methodologies


Projects in this thrust focus on the development of models, systems and methodologies needed to meet these goals. Thrust D projects, namely, sensing (D-1), regulatory control (D-2), informatics (D-3), supervisory control (D-4), and design (D-5) are specifically structured and integrated to accomplish these goals.

Key Accomplishments

Some of the key accomplishments of this thrust  are: 

 

  • Developed and fabricated novel sensing technologies for the implementation of online monitoring of critical process parameters. 

  • Designed and implementation of efficient control systems for continuous tablet manufacturing process (DC, WG, RC). Combined feed-forward/feed-back control strategies as well as hybrid MPC-PID approaches have been developed and evaluated. 

  • Developed techniques to diagnose and mitigate process faults. Can identify novel fault signatures for engineers. Can update fault database without disrupting process operation. 

  • Development of an integrated framework for the efficient incorporation of unit operation models using process simulation. 

  • Development of a real time optimization (RTO) framework to guide the optimal set points for the process unit control systems.  

  • Developed workflow based knowledge management system implemented on HUBzero platform and released through pharmaHUB.

  • A surrogate-based method has been developed to conduct feasibility analysis and control applications in dynamic systems. 

  • A reduced-order modeling methodology for incorporating information from detailed computationally expensive process modeling into a flow-sheet simulation.

  • Individual residence time distribution (RTD) models based on interpretation of experimental data have been developed for continuous powder handling unit operations. 

  • Development of sensitivity analysis approaches for determining the critical process parameters of the entire line without extensive experimentation. 

  • Development of a systematic methodology to determine the design space of the individual process operations and the integrated continuous line. 

  • Development of optimization approaches for determining the optimal operating conditions and design parameters of the integrated process system. 

Impact

During the entire duration of the C-SOPS but especially in the last 5 years Thrust D accomplishments in the areas of sensing, process integration, modeling, and optimization are of tremendous importance and have received great attention from the industrial mentors and the scientific community. 


Sensing using NIR and Raman spectroscopy has made possible the online measurement of critical process parameters. New sensing methodologies have been developed to further assist novel production routes.  Using the PAT tools the integrated continuous direct compaction tablet manufacturing process (feeders, blender, mill and tablet press) with the control platform for centralized automatic operation was demonstrated. 


For the first time we are able to demonstrate the value of flow-sheet modeling in: (a) identifying the critical process parameters; (b) design the most efficient control strategy for the individual units and for the entire process; (c) determine the range of conditions where the process remains valid (design space); and (d) identify the optimal operating parameters for the process to produce at the maximum efficiency. The use of advanced control strategies revealed the importance of closed loop production and integration of a data management capability. 

Future Plans

The work in the integration thrust will continue towards the development of novel sensing methodologies, the integration of these methodologies within the process production lines, the use of modeling to develop efficient control strategies, and the implementation of closed loop control for the optimal utilization of process capabilities. In particular we are targeting to: 

 

  • Increase sensing capabilities utilizing novel sensing technologies for online use. 

  • Integrate the PAT and control strategies for DC, DG, and WG lines for the production of solid based drugs. 

  • Integrate of PAT and control strategies for strip-film formation (test bed 2).  

  • To expand the center capabilities in process modeling and integration for all different production platforms.

  • To generalize the knowledge management tool to be utilized in continuous tablet manufacturing platform.