Machine-learning models estimate how long each task will take from measured productivity – and get more accurate with every completed task.
A constraint-optimisation algorithm puts every task on a suitable resource at the best moment: no overlaps, within the shift and after the preceding operations.
Receives tasks from StoreIT or any other system through standard messages and returns the finished schedule automatically.
In the warehouse and on the shop floor, work changes every hour: new orders arrive, people are absent, priorities shift. A spreadsheet schedule is out of date before it is finished, and the workload of people and machines stays uneven.
PlanIT does the calculation: for every new request it builds a schedule that respects all mandatory rules and puts the important work first.
In the 2026 tests PlanIT planned a request with 20,000 tasks for a warehouse with 150 workers in six areas – receiving, put-away, picking, packing, shipping and replenishment – and 13,200 dependencies between the tasks.
PlanIT does not replace the warehouse or ERP system but works alongside it: the customer's system stays responsible for the master data, PlanIT calculates the schedule.
Key figures: tasks planned today, tasks waiting, failed requests and deliveries, average task duration
Recent activity in real time
Resource Gantt chart – colour by priority, working time and predecessor → task links
Filter by resource and full-screen view
Task types for each company or warehouse
Resource pools per task type
Resources with daily capacity and pools in order of preference
Working time and unavailable periods for every resource
The data comes automatically from your system and can be checked in the client
Every request with its status – new, in process, planned or failed
The tasks of the request right below it
Double-click opens the request’s schedule – only its tasks and resources
Re-plan a failed request in one step
All tasks with quantity, priority, status, resource and planned time
Filters, sorting, saved views and export
Task details with its direct predecessors
The whole predecessor chain on a Gantt chart
Every incoming message – status, number of objects and errors
Request and response (JSON) side by side
Deliveries of the finished schedules: status, HTTP code, attempts and the receiver’s answer
Automatic retries and an archive of every message
A separate model for every company and task type
Active and archived model versions
Accuracy metrics (MAE, RMSE, median absolute error)
Training settings tuned automatically by a genetic algorithm
Productivity data – the actual start and end of every task, which the models learn from
Customers (invoice addresses) and their companies and warehouses
Message archive per customer
Users, roles and rights per function block
Menus, grids and forms configured without programming
Planning and training parameters
PlanIT helps wherever many tasks have to be shared between people and machines with different skills and shifts.
PlanIT combines proven libraries with code developed specifically for planning warehouse and production tasks.
We start with your real data and reach a working solution step by step.
Task types, resource pools, resources, working time and the planning requests. The format is described in the OpenAPI documentation with examples.
The absence is sent as an unavailable period, and PlanIT does not place tasks in it the next time it plans.
After the work is done, your system reports the actual start and end. PlanIT trains a model for each task type from them.
No. PlanIT works on its own with any system that can send and receive JSON messages. With StoreIT the connection is ready.
In the cloud or on your own server (Linux, PostgreSQL). Each customer’s data is kept separate.
In the tests a request with 20,000 tasks was planned in about one minute.
PlanIT was developed under project BG16RFPR001-1.001-0276-C01 "Development of an innovative cloud service for planning production and logistics tasks based on artificial intelligence at IFD-ENGINEERING JOINT VENTURE Ltd", funded by the "Competitiveness and Innovation in Enterprises" Programme 2021-2027, co-funded by the European Union through the European Regional Development Fund.
