object detection in computer vision

Objectives

To control coke size variation for energy cost reduction

object detection in computer vision

Industry Served

Steel Manufacturing

object detection in computer vision

Solution

Ripik Vision

Context

A formidable player in the global steel industry, this 1 million tonne steel plant focuses on producing both flat and long products through its Blast Furnace (BF) and Basic Oxygen Furnace (BFO). With a keen emphasis on cost competitiveness, the plant is acutely aware of the supercritical importance of maintaining economic viability for long-term survival in the competitive steel market.

Problem

Traditionally, coke sizing process was conducted manually with a sampling frequency of just 1/shift, effectively 3/day. Hence there was no track to find out what sort of coal is being fed throughout the shifts and how it is impacting the fuel rate and efficiency of BF. Blast Furnace (BF) operating with a fuel rate exceeding the benchmark, often exacerbated by an excess of feed due to improper input coke control. The critical factor in this scenario is the uncontrolled reaction rate of particle size, contributing to process variability. This infrequent and random sampling approach has resulted in increased costs, making it difficult for the client to survive in the cost competitive industry of Steel manufacturing.

img

The Solution

Ripik.AI focuses on reducing operational variability through optimal process control. We implemented our patented Cognitive and Vision AI based solution Ripik Vision, capable of particle differentiation based on color, moisture and size. With the plant team interaction and first-hand process control, we swiftly delved into comprehending their requirements.

  • Real-time coke sizing by Ripik Vision increased their coke sampling from 3/day to 4000/day for an efficient input feed control.
  • It helped the team to pinpoint the anomalies contributing to the increased fuel rate in the furnace.
  • Also, the solution predicted slag chemistry and hot metal silicon, scrap quality in Electric Arc Furnaces (EAF).

This approach empowered operators to optimize fuel efficiency, enhance furnace performance, and contribute to more efficient and sustainable operations.

The Impact

In the first few weeks of deployment, our client saw

  • A significant 3% reduction in fuel use, which was ascribed to precise Coke size control and uniformity.
  • The integration of smart alerts, leveraging video feeds, and the capability for post-mortem analytics further enhanced their operational efficiency and productivity.
img

“Ripik.AI understands the complex blast furnace process and their knowledge in the field of data science has been invaluable in helping us optimise our plant operations and improve overall efficiency.”

Chief Digital Officer

Explore how Ripik Vision can transform your steel manufacturing operations.