Measure how closely your simulation matches the real machine.
We compare your model’s motion with measurements from the real unit, fit the effects agreed in scope, and report the remaining error. You can use the report to judge whether the model suits its intended task.
Compare commands and measured response.
We use recorded runs to fit agreed effects such as friction, damping and controller lag. The report identifies what was fitted, what remains approximate and the measured error.
A stage is one round of measurement and fitting over an agreed operating range. Each stage has a fixed scope and price and ends in a report. Further fitting is a separate decision.
The assessment uses your existing model or a separately scoped asset build. The report includes fitted parameters and identifies the simulator and model version tested.
- The command profiles the real unit and the twin were driven with
- Tracking error, real against twin, as figures and plots
- What was fitted from the logs, and what was left ideal
- The conditions each run was measured under
- Where each result was measured — our bench or your site
- What the model is not suitable for, stated plainly
Our rover comparison found a steering mismatch.
In a recorded arc test, the rover converted 0.34 of commanded yaw into rotation; the model converted 0.04 to 0.14. This identifies a substantial steering mismatch. Reliable straight-line driving and transfer of controller performance to the physical rover remain unproven.
Recorded yaw response
Fraction of commanded yaw converted to rotation; higher means more rotation for the same command.
What the first stage needs from you.
One shipped unit
A unit to measure on our bench or at your site. Installed machines are measured in their operating setting.
A way to command and log it
A driver or your existing tooling that can command repeatable motion and record the response with timestamps.
What the twin is for
The intended task and operating range, so we can agree which effects need fitting.
Start with one machine and operating range.
The assessment needs an existing model and a way to command and measure the physical machine. We agree the operating range, test conditions and report before starting. Model changes, further fitting and on-site measurements are scoped separately.
If measurements are needed at your site, we agree access, travel and the test schedule in the proposal.
Before we discuss your project.
What tolerance will you hit?
Achievable accuracy depends on the machine and the available measurements. The first stage establishes the error and includes the fitting agreed in scope. Its report helps you decide whether further work is useful.
Can you validate a twin we already own?
Yes. We compare your existing model with the hardware using the same command profiles. We agree the assessment scope before starting.
What if the twin turns out to be badly wrong?
The report records the mismatch, any identified causes, unresolved questions and limits on the model’s use. Further fitting is quoted separately; the assessment does not guarantee that a model will suit every task.
Does the report have to be shareable with our customers?
No. You decide who receives the report. Nothing about your product is published here without your written permission.
We run a simulator you have not mentioned.
The quote states where we will build and test, what can be transferred, and what adaptation is needed in your simulator.
Tell us about the machine and the model.
Start with a free twenty-minute scoping call. We’ll review the task and available information, then propose a scope and price. If a paid assessment is needed first, we’ll agree that with you before proceeding.
Or email hello@drimkoe.com
For procurement: the capability statement — competencies, past performance, scope and engagement terms in a printable document.