Kinta Updates: New Product Features

Since raising our $5.5M Series A in January, we have been working hard at Kinta on making progress at further enabling the factory of the future. Rapid hiring and investment in top-notch engineering talent has opened up the possibilities for our product’s development. In this blog, we’re pulling back the curtain on our algorithm, our near-term product roadmap and going deeper on our newest features. We will talk about KPI diffs, scenario planning and purchase justifications.

In our first blog post, we wrote about stages of data literacy, and how AI software is pivotal in helping manufacturers leverage data for informed decision making. Kinta’s algorithmic scheduling tool empowers planners to utilize all their factory’s data.

It is well established in operations research that finding the global optimum of a production plan -- the undisputed very best solution -- can be incredibly time intensive and impractical. In even very simple factories, there are more possible production plans than atoms in the universe. Factory environments change faster than global optimization can possibly compute. Therefore, at Kinta, we utilize AI techniques to locally optimize schedules and enable manufacturers to create data-driven schedules. 

Beneath the surface, the key technical components at work for algorithmic optimization are Reinforcement Learning and Integer Linear Programming. The interweaving of these techniques enables us to bypass training on historical data, and allows greater speed and flexibility in calculating and recalculating schedules. Finding practical optimal solutions allows planners to quickly create an efficient plan, and allows them to get back to managing their shop floors.

Leveraging this AI-powered scheduling platform with our clients has led to dramatic improvements to core operational KPIs in some of the largest factories. From decreasing changeover times by 21 percent to increasing on-time delivery by 14 percent, we’ve been able to generate cost savings of millions of dollars. While keeping track of KPI and analytics have always been key to the Kinta user experience, we’ve now made it even easier to track KPI wins on a day-by-day basis through our new feature: KPI diffs.

Traditionally, tracking KPIs was a highly manual process: employees would pull together reports once a week, and if specific data was needed, somebody had to check machine statuses on the shop floor. With the right integrations, software tools like Kinta can collect the important data, such as changeover times or makespans, and provide raw data for analytics, or even directly generate graphs.

To track changes over time, or even compare metrics across different iterations of the same schedule, planners would traditionally have to manually compare the relevant numbers between those iterations. Now, with KPI diffs, those comparisons are calculated instantly within Kinta under the KPI summary tab. Each relevant KPI on the list, which can be customized based on the user’s preferences, will be shown with the change in its value as the planner edits the schedule.

Having KPI diffs to track metrics such as utilization or on-time delivery makes it even easier for planners to stay close to their factory’s data and analytics. Planners will no longer have to estimate what the best schedule is, but they can actively choose the optimal schedules and justify their choices by evaluating the KPI differences between schedules. Making data-driven decisions will help actions become justifiable and more transparent so planners can consistently win the day.

KPI diffs is one new feature that is particularly powerful when it comes to scenario planning. Another recent update that will greatly improve scenario planning is enabling planners to easily draft schedules for different scenarios for comparison. When the planner has sufficient space to test out scenarios, they can simply choose which best fits their facility’s and customers’ needs that day. 

 Within each draft schedule, Kinta is also developing the capability to edit data. This means planners can update sales due dates or order priorities, but the implications of this feature go beyond just production planning. By being able to edit the data to add in machines, planners will now be able to more easily perform bottleneck analysis and capital investment modeling on specific lines to find out how much adding or removing a machine would affect KPIs. 

In short, Kinta’s recent product developments will help planners get the most out of their resources and factories in three new ways: KPI diffs, scenario planning and purchase justifications. To learn more about Kinta, our algorithm, or what’s in store for the future, don’t hesitate to reach out to our team at [email protected] Or, if you want to see our product in action click here to schedule a demo today.