Background Image

Engineering Solutions

Domain Agnostic algorithm to meet challenges in manufacturing.

Our solutions involve rigorous computational techniques that can be deployed to large chemical processes and also individual process sub-systems. Our strength lies in data-reconciliation, gross error detection, modeling and simulation, early stage process-design, control and optimization.


Domain agnostic algorithms of a solution across multiple manufacturing sectors. Text / sentiment analytics.

Process industry

  • Data reconciliation & optimization
  • Hydraulic modeling and surge analysis
  • Implementation of expert systems
  • Enterprise benchmarking & data analytics
  • Controller performance monitoring & diagnosis
  • Data acquisition through image processing

Oil & Gas

  • Advanced real-time refinery monitoring & optimization (in collaboration with PSE & BPCL)
  • Assessment of feed to crude distillation unit
  • Soft sensing studies

Foundry solutions

  • Casting defect prediction
  • Optimal sand-plant operational state
  • Additive blending models for reaching optimal & stabilized operation of the sand plant

Food industry

  • Automation systems for waste minimization in the food industry
  • Supply chain solutions


  • Data driven models for predictive analysis and early warning systems based on image processing
  • Condition monitoring of PCMM direct compression
  • Proof of concept studies for use of 3D-printing in pharma


  • Model-on-demand framework for modeling fuel cells
  • Efficient closed form solutions for optimal load sharing strategy in fuel cell networks
  • Rapid health monitoring & diagnosis for fuel cell networks

Financial analytics

  • Mean reversion in pair trading
  • Data mining - financial message boards
  • Improving momentum based trading strategies
  • Audit data misstatement identification
  • Journal entry bifurcation
  • Insider sentiment score for US stocks
  • Transaction streams network modeling
  • Market neutral portfolios


  • Sea navigation analytics
  • Multivariate steady state detection & data reconciliation for thermal power plants
  • Programmatic identification of optimal coil design for coolant in air-conditioning systems

Skill Sets

Typical manufacturing challenges addressed by our solutions.

  • Customer Relationship management instead of financial analytics
  • Training to be removed
  • Predictive Maintenance
  • Data integrity assessment and pattern deduction.


Increased competition, stringent environmental regulations and reduced profit margins
are hurdles to sustenance and growth – harsh realities that every business
has to confront and mitigate in their mission to remain successful.

In this scenario, reducing wastages and operating the plant as tightly and optimally as possible is of critical importance. In an operating plant, sources of losses and opportunities for process improvement are hard to identify. Since one can't fix what is not understood, it is business as usual for many plants. Gyan Data provides services for process improvement through modeling and simulation, data analytics and our in-house tools.

Our areas of expertise include :

  • Power plant analysis
  • Pipeline network design
  • Surge analysis
  • Battery management
  • Scheduling for batch plants
  • Scheduling of multiproduct pipelines
  • Crude pre-heat-train optimization
  • Uploading operations
  • Controller performance assessment


Packaged algorithms that address specific challenges faced by the manufacturing industries.

  • Refinery crude properties assessor
  • Sensor and Equipment Audit Tool (SEAT) Modern process industries are equipped with data acquisition systems that routinely and automatically gather data from hundreds or thousands of sensors in a few seconds. This sensor data is used for process improvement through optimization and process equipment health monitoring. It is important that the integrity of the sensor data is verified before it is used in other applications. The Gyan Data tool can reconcile sensor data, identify gross errors in sensors and use the reconciled sensor data in developing key performance indicators for individual equipment. This can help plant personnel avoid routine maintenance schedules and target maintenance where there is a clear and present need. Such an approach will result in reduced maintenance costs and improved equipment performance. In keeping with our guiding principles, the Gyan Data tool can be used at a single equipment level with any level of existing instrumentation, or any subsection of plant that is thought to be of critical importance.
  • Controller Performance Assessment Tool (CPAT) Gyan Data Controller Audit Tool is geared towards diagnosing the cause for oscillations in closed-loop control systems. Oscillations in control loops lead to millions of dollars in additional operating expense for manufacturing plants yearly. Additionally, an oscillatory control loop limits the maximum quality achievable by the system. The tool can be used for single loops or multiple loops. The Gyan Data approach can detect oscillations in control loops, and describe the root cause(s) for oscillations without requiring costly off-line time for the system. The additional ability of this methodology to differentiate between external causes, stiction, and poor tuning allows system managers to intelligently schedule maintenance and tuning efforts.
  • Steady State and Dynamic Performance Audit and Optimisation Tool
    Any attempt at process improvement starts with operational data which has to be analyzed and interpreted to extract valuable information about the process. Gyan Data's tools help personnel to understand process performance improvements based on not only just operational data (traditional data analytics) but also a judicious combination of data and first principles modeling (hybrid analysis), focused on individual units or subsections of a plant. The deliverables for the customer include recommendations on changes to operational strategies with quantified post implementation benefits, recommendations for hardware changes that can improve performance with the associated cost-benefit analysis, and development of customized algorithms for online implementation within the existing IT framework in the plant.
  • Matlab, R, Python

Case Studies

Use cases that we have worked on for various industries verticals.