Can it start from the customer’s drawing?

Rebuild the warehouse from the customer’s PDF drawing as a scaled underlay, racking, pick zones, benches and dock doors as parts with ids, and record what the drawing did not give as an assumption.

How are robot counts compared under the same demand?

Run the peak with two, three or four robots, and with the packing benches moved, under the same arrivals and paired seeds; read throughput, the wait at each pick-and-drop point and utilisation, with the difference and its interval.

Does it model collisions or real robot behaviour?

Results come from a simplified, deterministic model: robots are circles moving at constant speed, and the same input always gives the same result. It does not model collisions or avoidance between robots or people (only capacities and queues), crowd physics, optimised lift group control (lift cars follow simple fixed rules), turning, or cleaning quality. Batteries, when a scenario gives them, drain in a straight line and charge at a constant rate. Dimensions read from photos and drawings are assumptions to check against measurements. It is for exploring options, not a calibrated prediction of a real site.

For warehouse robot integrators

A customer sends a warehouse drawing and a peak throughput and asks how many robots, and why. You need a model that matches the drawing, runs with two, three and four robots and one layout change under the same demand, and a pack with the routes, the chargers and the numbers behind them.

Open the warehouse from the robot fleet proposal, without signing in Connect your AI

What Laydyne does for it

Rebuild the warehouse from the customer’s PDF drawing as a scaled underlay, racking, pick zones, benches and dock doors as parts with ids, and record what the drawing did not give as an assumption.
Read the feature page: Build a 2D plan and a 3D model from a drawing, photos or a brief
Run the peak with two, three or four robots, and with the packing benches moved, under the same arrivals and paired seeds; read throughput, the wait at each pick-and-drop point and utilisation, with the difference and its interval.
Read the feature page: Simulate how people, goods and robots move and wait
Show the existing layout and the proposed one from the same viewpoint, with the parts that moved listed.
Read the feature page: Put two layouts in the same view and read the difference in numbers
Issue the proposal drawing with the robot routes and the chargers, and the route sheet with the first seed’s complete traces.
Read the feature page: Issue A3 plans, quantity sheets and a PDF report from the saved version
Hand over the comparison as CSV and JSON beside the editable project, so the customer’s own AI can reopen the options and the runs.
Read the feature page: Keep the project as a file you own, and hand the model on

Open a recorded example

What the published records produced

The fictional 60 × 40 m warehouse study uses 20 matched seeds. Four robots at the existing benches complete 102.38 jobs an hour. Three robots at the moved benches complete 112.06. The paired difference is 9.68 jobs an hour, with a 95% interval of 7.30 to 12.06. These counts include charging jobs and do not establish a guaranteed tote throughput or a minimum fleet. In the separate 3PL study, six pickers with three pack stations leave 0.3 orders missing the cut-off on average over three seeds. The baseline leaves 30.7.

These recorded projects use synthetic inputs. Photos and drawings show what was produced; site accuracy, business effects and photorealistic reconstruction are not established.

What it does not do

Results come from a simplified, deterministic model: robots are circles moving at constant speed, and the same input always gives the same result. It does not model collisions or avoidance between robots or people (only capacities and queues), crowd physics, optimised lift group control (lift cars follow simple fixed rules), turning, or cleaning quality. Batteries, when a scenario gives them, drain in a straight line and charge at a constant rate. Dimensions read from photos and drawings are assumptions to check against measurements. It is for exploring options, not a calibrated prediction of a real site.
Collisions, avoidance, following, passing and re-routing between agents (robots or people): only capacities and FIFO queues are computed.
Agreement with real robots or real sites: determinism guarantees reproducibility, not field accuracy; calibrate separately.
Each picker takes one order per cart: batching, aisle congestion, picking errors, breaks and absence are not modelled. The goods lift has one car with a capacity and travel times, not a manufacturer’s dispatch logic. Labour cost and carrier capacity are not calculated.

Price

Building and editing the model with your own AI over MCP, route checks, saving in this browser and JSON import and export are free, with no count limit; a free account is needed only to create an MCP link (3 per account). Running simulations, the built-in AI and cloud saving need Pro ($29 a month or $278.40 a year) or Max ($89 a month or $854.40 a year). Prices are in US dollars before tax; your AI provider bills its own usage separately.

What is free, and what Pro and Max add

Common questions

Can it start from the customer’s drawing?: Rebuild the warehouse from the customer’s PDF drawing as a scaled underlay, racking, pick zones, benches and dock doors as parts with ids, and record what the drawing did not give as an assumption.
How are robot counts compared under the same demand?: Run the peak with two, three or four robots, and with the packing benches moved, under the same arrivals and paired seeds; read throughput, the wait at each pick-and-drop point and utilisation, with the difference and its interval.
Does it model collisions or real robot behaviour?: Results come from a simplified, deterministic model: robots are circles moving at constant speed, and the same input always gives the same result. It does not model collisions or avoidance between robots or people (only capacities and queues), crowd physics, optimised lift group control (lift cars follow simple fixed rules), turning, or cleaning quality. Batteries, when a scenario gives them, drain in a straight line and charge at a constant rate. Dimensions read from photos and drawings are assumptions to check against measurements. It is for exploring options, not a calibrated prediction of a real site.
Is a simulation run free?: Building and editing the model with your own AI over MCP, route checks, saving in this browser and JSON import and export are free, with no count limit; a free account is needed only to create an MCP link (3 per account). Running simulations, the built-in AI and cloud saving need Pro ($29 a month or $278.40 a year) or Max ($89 a month or $854.40 a year). Prices are in US dollars before tax; your AI provider bills its own usage separately.