At Kinta, we pair cutting edge Operations Research with Artificial Intelligence to deliver a complex, yet flexible solution capable of driving value in many facets of your manufacturing business. Read our case studies below for real-world use cases where production planning optimization has unlocked value for our clients.
For a Stanley Black & Decker fasteners plant in Stanfield, North Carolina, Kinta helped get more value out of each machine. The Stanfield team was able to allocate plant resources such as machines, tools and labor more effectively with AI-driven production schedules, resulting in double-digit improvements to queue times, throughput and on-time customer delivery.
At Diebold-Nixdorf, Kinta’s algorithms were able to decrease inventory holding costs and reduce throughput variance by leveraging material availability simulations. Planners were given advance notice of potential issues, so they could automatically schedule and reschedule accordingly. This resulted in more efficient sequencing across departments and over a million in annual recurring savings.
A large cement manufacturer in Europe was able to reduce their carbon footprint and energy costs by following Kinta’s production schedules. By optimizing the cement routings with Kinta, the client saw reduced electricity consumption, reduced finished goods inventory and hundreds of thousands of dollars in annual cost savings.
Kinta is designed for manufacturers of all sizes, in all industries. Our tool excels in facilities where there is at least a medium level of data availability and operational complexity. See our case studies above for examples of successful implementations and how you could also meet delivery dates, reduce changeovers and overtime, and maintain lean production methodology.
To get started we’ll need a few basic data tables, such as a list of machines, routings and cycle times, as well as a demand signal. During implementation, we’ll meet regularly with an IT stakeholder to oversee proper data connections, and a factory stakeholder who can describe your operations and business goals.
No, Kinta only needs current factory data to build a digital model of your factory. Just like modern Chess or Go engines, the algorithm optimizes a plan during a fast simulation of a single planning period.
Kinta does not train on a historical training set at all, since it uses Reinforcement Learning rather than Supervised Machine Learning.
Kinta is priced as a per-factory license, with bulk discounts for additional factories. By eliminating per-seat pricing, everybody in the factory can access Kinta, so the platform can become a frictionless source of truth across departments.
We take data security very seriously at Kinta and ensure our security protocols are always up to date. We use AWS as our hosting service and adhere to strict SOC2 data security standards.