We are a Flatworld Solutions company.
Smart factories does not warrant invasive infrastructural changes or redesigning them completely. Analytics and machine learning models work with a range of data, even hand-written notes, and provide a scope of improvements that can increase productivity by atleast 15-20%. People, products and production facilities are fully connected through the Internet of Things (IoT). Sensors collect vast amounts of data, and that information is then interpreted by Artificial Intelligence (AI). This optimizes the production line and creates synergies across the facility.
Leveraging the power of AI for an industry that is characterized by assets-intensiveness and high equipment costs is usually met with speculation owing to a slower ROI (when compared with other discrete manufacturing units). However, AI and Industrial IoT can realize benefits from the smallest value to broader business outcomes. Quadratyx offers a spectrum of solutions that range from simple-yet-intensive analytics with existing data collection mechanisms to complex, holistic plant management solutions.
Increase in Productivity
Losses Identified
Decrease in Breakdown
Upto
Increased Case Acceptance % & Revenue
One of India’s largest steel manufacturers, that specializes in a range of pipes; identified losses and determined that there was an opportunity to improve productivity by 15-20%.
Quadratyx deployed a customized predictive maintenance systems at this plant, that utilized historical and current data, already captured at the plant. The array of ML models were able to identify unaccounted losses, identified upcoming failures and yielded an increase in productivity that was over three time the expected value. Inherent issues with data collection methods were also rectified.
A leading Dairy and Foodstuff company in the Middle East. (Revenue ~ $495Mn yearly, 35k customers, 24 depots). The challenge for the company was to manage 500 routes daily -which led to the problem of streamlining delivery activities and optimizing daily routes. The company wanted to centralize routing and scheduling process of the trucks.
By leveraging our in-house ML based proprietary solution, ARRO (Adaptive Road Route Optimization), the firm optimized its delivery plan after taking into cognizant all the constraints such as fleet, location, time and other domain specific constraints.
Save fuel by reducing the
Distance Driven
by
27%
16%
Reduction in number of routes and
Improve vehicle utilization
24%
Reduction
in
Time of Travel
Achieved
Overall Accuracy
of
92%
Improved
Delivery Efficiency
&
Maximize Value
of visit to the store
Maximize
Maximize Personnel Productivity
A full-service distributor of global consumer brands (Revenue ~ 50M USD yearly, presence in 9 countries covering 22 markets) delivered products on a periodic basis to each of the outlets. Often product recommendations to retail outlets is based on salesman intuition. Thus, this ends up in expectation mismatch.
Quadratyx PRE-B (ML solution for recommending outlet purchase behavior) was deployed to predict the correct quantities of various goods that the delivery vans must be stocked with.
Incentivizing good behavior and ensuring that benefits reach to dealers in the network is an important aspect of channel visibility. Many sectors with huge distributor dealer channels face this problem of lack of transparency in this operational model. Our client was facing similar challenge.
Quadratyx’s customized solution was deployed for one of India’s largest steel manufacturers. The AI system for channel visibility and dealer retention allows the huge group of dealers to be rated and compared. It also analyses and predicts dealer performance in the future based on data submitted in the past. Analysis and rating is done for specific and multiple KPIs.
The system also allows for sentiment analysis through text analytics of comments and reviews given by end customers.
Interactive Dashboard
Rate and compare Performance
Sentiment Analysis
Competitor Analysis
At times flaws in products get unnoticed, even if the inspector is very experienced. However, machines equipped with cameras can detect even the smallest defects. Once the images are recognized, semi-supervised ML is the most effective technique to classify images into failure classes. The main benefit is cost reduction, , helps eliminate waste and optimize productivity for both supplier and customer.
Sharply fluctuating demand, supply of raw materials and uncertain
prices are few challenges in the supply chain of a production unit. For these tasks, techniques such as time-series analysis, probabilistic modeling (Markov and Bayesian models) as well as simulations (e.g., Monte-Carlo simulation) are most commonly used to reduce inventory planning time, minimize inventory cost, optimize repairments, and find optimal reorder points.
‘What’ to produce and ‘when’ is an important decision for any mill, and it is particularly critical when one of your most important inputs is electrical energy. The optimization models maximize energy consumption at off-peak times and thus minimize energy costs.
ML models use macroeconomic data, historical demand data, process statistics and transactional data to predict future demand. Scrap metal is a key raw material in most steel plants and its availability is not always certain. AI models can develop a “scrap index” and use hedging approaches while buying scrap steel.
© 2014-24 Quadratic Insights Pvt. Ltd. All rights reserved
