Logistics and kitting
Operators are guided to the correct bin by AR projection or DMX pick-to-light, while the 3D sensor validates each pick. The number of items to pick is shown and logged.
How it works
From MES job to production data.
Every production environment relies on one flow: design, execution and improvement. Arkite connects all three, from the engineer's desk to the workbench and back to the data.
Last updated
The operational flow
Five stages, one platform.
- 01
Jobs arrive from the MES
Jobs are created in the MES and pushed to the Arkite server through the REST API, including product variants and operator assignments.
- 02
Engineers build and adapt
Process engineers translate standard operating procedures into Arkite jobs, teach detections and add projections, so instructions are practical for the physical workstation.
- 03
Deployment to the right station
Once approved, jobs are deployed from the server to the correct workstations and are immediately available across the factory.
- 04
Guided, validated execution
The operator logs in, the workstation receives the correct job and variant, and each step is projected on the workbench and validated as it is performed.
- 05
Data back to the server
Every workstation feeds cycle times, step durations, errors and torque values back to the Arkite server for analysis in Arkite dashboards or BI tools.
One job, from MES to data.
Video placeholder Explainer sequence following one job, 90 seconds: the order arrives from the MES, an engineer adapts the job in the editor and publishes it, an operator scans a badge, the workstation loads the job and variant, steps are projected and validated by the 3D sensor, and the cycle time appears on the dashboard. Screen recordings combined with workstation footage; demo product only.
Engineering
Work instructions built without code.
Arkite is built on a no-code workflow engine. Engineers define jobs, teach the detections for bins, parts, tools and virtual buttons, and structure the job in numbered steps and sub-assemblies. Projections and text instructions are generated automatically when a detection is taught.
Any PC connected to the server can act as an editor, so projects are prepared at the engineer's desk without stopping production and can be run for up to 15 minutes for testing. Since software version 2026.2, engineers can also generate a job with AI: from a written description of the assembly, the software generates the job, job options, materials, detections and job steps, and the engineer reviews and finalizes the result. The same optional AI integration translates all instructions and texts from the default language.
- Conditions, loops, variants and communication with external systems
- Text, images, video and icons per step
- Version history with Draft, Review and Production status
- Optional approval workflow before release to Production
- Optional AI job generation and translation, configured with your own key
At the workstation
The right job, the right variant, the right level of guidance.
Operators log in by scanning their badge, which also sets their experience level. Through the MES integration, the workstation retrieves the correct job and product variant automatically, which removes manual selection errors.
Instructions are projected on the workbench and shown on the operator screen, step by step. Novice operators receive more guidance; experienced operators work with only what they need. With Arkite VALIDATE, the 3D sensor verifies each pick and assembly action and the workflow continues automatically after a correct action.
Across the shop floor
One platform for every manual process.
Assembly
Step-by-step guidance across products and sub-assemblies, with variant logic, smart tool control and real-time validation of each action.
Packaging
Clear digital guidance for consistent packing, including projections driven by a stacking algorithm from an external system.
The data loop
Every workstation reports back.
Each workstation continuously sends structured data to the Arkite server: cycle times against expected cycle times, step durations, errors, torque values from integrated smart tools, items picked and traceability records per serialized unit.
The server becomes one central source of real-time production data from manual operations. Built-in dashboards show workstation status, job output, cycle times and live step durations. The same data can be used in Power BI or any other BI platform to find bottlenecks, optimize cycle and takt times and reduce scrap and rework.
Design, execution and improvement: one flow, one platform, one source of production data.
Related
Arkite AI
AI job generation, AI translation and VISION AI detections, with data under your control.
GUIDE, VALIDATE, VISION
Choose the validation depth each workstation requires.
Integrations
How Arkite connects to MES, ERP, PLC and smart tools.
Production intelligence
What data Arkite captures and how it is used.
Picking and kitting
Pick-by-projection and pick-to-light with 3D validation.
Walk through the flow on your own product.
In a demonstration, we map the Arkite flow to your MES landscape, workstations and variants.