Engineering validation
Case Studies & Engineering Notes
Customer case studies are published only for actual projects, with the customer's written permission. Until the first is cleared, this page sets out how DOKTEK validates an automation project, including an anonymized eggplant stem cutting example.
Publication status
No case study is published yet. DOKTEK publishes only actual projects, with the customer's written permission, in the customer's own numbers. Until the first one is cleared, this page shows the template every case study will follow and how a project is validated — so a prospective customer knows exactly what will and will not be claimed.
Engineering validation notes
What every food automation project has to prove.
The questions below are settled on the processor's own product before a machine is quoted. They apply to eggplant stem cutting and to every other irregular food product DOKTEK evaluates.
- 01
Real product trials
A machine is validated on the processor's own product, not on hand-picked samples. Trials use fruit across the full size and quality range the plant actually receives.
- 02
Product variability observations
Size, curvature, calyx shape, stem angle and surface condition are recorded before any configuration is proposed, because they decide what the machine has to handle.
- 03
Machine vision challenges
Glossy or dark surfaces, green parts against the belt, random orientation and touching products are evaluated on the processor's images, under the lighting the machine will use.
- 04
Feeding and orientation challenges
Vision handles random orientation; it cannot handle stacked or overlapping product. Single-layer infeed and spacing are designed together with the upstream line.
- 05
Cutting consistency requirements
A correct cut is defined before the trial: acceptable residue, flesh loss, and how missed or over-cut products are counted.
- 06
What must be confirmed during testing
The handled size range, cut quality per category, the feed rhythm at which quality holds, out-of-range handling, cleaning access and documentation.
- 07
Why every project needs product trials
Variety, growing conditions and upstream handling change what reaches the machine. That is why no fixed capacity or accuracy figure is published before a trial.
Template
What every DOKTEK case study contains.
- 01
Customer / region
Named with permission, or an anonymous label such as 'European vegetable processor'.
- 02
Process challenge
The manual operation and why it resisted automation.
- 03
Manual method
How the plant did it before, with the measured rate and reject figures the plant supplied.
- 04
Target
The specification the project had to meet.
- 05
DOKTEK solution
The machine or station proposed, with its status at the time.
- 06
Engineering approach
Sensing, software and motion decisions, and why.
- 07
Machine configuration
What was built for this product and line.
- 08
Validation
The trial on the customer's product, how it was measured and what was recorded.
- 09
Result
Only measured outcomes, in the customer's numbers.
- 10
Video
DOKTEK's own footage of the machine on the customer's product, where permitted.
- 11
Next steps
What follows: further stations, retrofit, or nothing.
Validation path
From sample to documented result.
This is the path a project follows before it can become a case study. It is also the reason the machines pages carry honest statuses instead of catalogue promises.
- Step 1
Samples and video
The customer sends product across its full quality range and a short video of the manual method.
- Step 2
Engineering review
DOKTEK identifies the operation, judges feasibility and proposes what to validate.
- Step 3
Trial on the customer's product
Real leaves, real fruit, real recipe — measured against the customer's criteria and recorded.
- Step 4
Documented result
Measured outcomes in the customer's numbers, with the configuration that produced them.
Evaluate your process with DOKTEK engineering.
Send a short video of the manual process and a few product photos. An engineer replies with what can be evaluated before any quotation.