For photometric and lighting layout consultants
Spend your hours on judgement, not on moving poles. Runnax Engine searches pole layouts for a parking lot in about a minute, keeps the fewest poles that meet your targets, and puts every input on record so you can rebuild and sign off in AGi32.
Where it fits in your work
- You set the job: lot outline and drive aisle, where poles may stand, the IES files, a pole-height range, and your minimum and Max:Min targets.
- The engine searches: regular arrays of up to 16 poles in up to 3 rows, only inside the allowed areas, and keeps the fewest poles that pass, or says why nothing does.
- You check it: pole table, horizontal and vertical point grids, isolux sheet and a summary with every input, ready to rebuild in AGi32.
- Before a file goes out: an AGi32 second opinion recomputes every zone of your file and flags what it assumes, such as a light loss factor of 1.0.
How close it is to AGi32
| Comparison | Result |
|---|---|
| Real parking lot, 271 points, vs an AGi32 direct-light run | Minimum, average and maximum within 0.5 % |
| Same lot, largest difference at any point | 0.32 % of the average |
| Three real AGi32 parking-lot files, recomputed from the file | Zone averages within 0.06 % |
Every comparison and what is not validated yet.
Current scope
- Parking lots: regular arrays, one mounting height per lot, one parking zone per run, straight aisles.
- Direct light only; no obstructions or shadowing. Simulations, not field measurements.
- It checks the targets you set; choosing them and signing off compliance stay with you.
Your clients' files
Used for your run only and never shared with another customer; originals deleted 30 days after the run; no case study without written permission. Free runs may help improve the engine with de-identified data, and you can opt out. How we use data.
Questions
Does it replace AGi32?
No. It replaces the hours of moving poles and re-running a parking lot until it passes. You set the pole areas, targets and luminaires, the engine finds the layout, and you check and sign off in AGi32 with every input on record.
How accurate is it against AGi32?
On a real parking lot, the engine's minimum, average and maximum are within 0.5 % of an AGi32 direct-light run over 271 points. Recomputing three real AGi32 parking-lot files, zone averages agreed within 0.06 %.
Will my client's project be shared?
No. Files are used for your run, never shared with another customer, and the originals are deleted 30 days after the run. No case study without written permission.
Can I call it from my own tools?
A REST API, an MCP server and a CLI ship together with the API release; they are not public yet. Until then, early-access runs go through us.