What RD testing begs for...

daryl adams

Premium Member
Premium Member
General User
Newly Registered
Nov 3, 2018
8
0
#1
There are so many variables across test beds in the area of RD test. Most testing, although well intentioned, involves an exact state of traffic/ signal that can never be repeated. Not really fair. Wouldn't it be nice if we had a testbed that had a "set" track with a "known", challenging set of problems/ issues for an RD to address/ resolve? Static distance, signal, obstructions, noise, other environmental factors. The holy grail of test beds. A test bed from which not only different detectors could be challenged/ tested, but different versions of firmware could be tested. No one provides detailed tests across firmware versions of all the available platforms. We "assume" no changes have been made. We "know" however that variations in function effect other daemons/ functions on a fixed hardware platform with fixed resources. In the end, aren't we really just testing firmware? Doesn't it make sense to test across a fixed test bed?
 

Deacon

TXCTG
Premium Member
Advanced User
Intermediate User
Newly Registered
Nov 13, 2016
7,427
91
Hill Country, TX
#2
If you have an old airport you’d like to donate for the purpose, I’ll get working on it.
 

Heywood

Learning Something New, Still Dying Stupid
Premium Member
Advanced User
Intermediate User
Beginner User
General User
Newly Registered
May 12, 2016
2,199
24
Alberta Canada
#3
Bonneville Salt Flats is all yours. Enjoy lol.
 

CJR238

-CMS Article Admin-
Administrator
Premium Member
Advanced User
Intermediate User
Beginner User
General User
Newly Registered
Oct 30, 2010
26,219
27
CT. MZDA 3. ALP, V1/RedLine/Waze
#4
There are so many variables across test beds in the area of RD test. Most testing, although well intentioned, involves an exact state of traffic/ signal that can never be repeated. Not really fair. Wouldn't it be nice if we had a testbed that had a "set" track with a "known", challenging set of problems/ issues for an RD to address/ resolve? Static distance, signal, obstructions, noise, other environmental factors. The holy grail of test beds. A test bed from which not only different detectors could be challenged/ tested, but different versions of firmware could be tested. No one provides detailed tests across firmware versions of all the available platforms. We "assume" no changes have been made. We "know" however that variations in function effect other daemons/ functions on a fixed hardware platform with fixed resources. In the end, aren't we really just testing firmware? Doesn't it make sense to test across a fixed test bed?
That's why we do multiple runs on a predetermined low traffic course. Some test courses also test different aspects of a RD. Like speed, long range, off axis...

Generally different versions of firmware do not effect performance and more times than not its mostly feature based differences.

Even Airports and flat salt beads would not provide as accurate testing as it does with real roads.

Good thread for general testing guidelines: www.rdforum.org: How to test your RD for general use
 

HeartDoctor

PSL + 5 never more;)
Advanced User
Intermediate User
Beginner User
General User
Newly Registered
May 30, 2016
1,738
0
SE Florida
#5
Your testtbed better be in an climate controlled environment because humidity, atmospheric pressure, suspended dust, pollen, air temperature, etc. might all impact results.
 

Heywood

Learning Something New, Still Dying Stupid
Premium Member
Advanced User
Intermediate User
Beginner User
General User
Newly Registered
May 12, 2016
2,199
24
Alberta Canada
#6
Your testtbed better be in an climate controlled environment because humidity, atmospheric pressure, suspended dust, pollen, air temperature, etc. might all impact results.
Probably the best reason a testbed scenario is unrealistic. You can throw temperature in there as well.

Even the placement of a radar emitting devise. Such as inside a warm vehicle, or an antenna mounted outside changes some things.

The best is constant testing in different environments and taking an average over multiple runs as we do now.

We have wild DA’s (Density of the Air or atmosphere) where I am. If you race you know what I’m talking about.

I can go out Friday and get a DA of 2200 ft. Which is correct for my location. I’ll go out Monday and the DA will be equal to 4400 ft above sea level.

I have to use the Dragtime app to bring the results down to sea level to see if I made any gains.

If we see results in about’s 300 ft of each other over various tests across the continent, i would consider them a good representation of what each detector is capable of.

The other issue is the capability of each detector. We do know that there are the occasional hot copies that comes out and squew’s the results between 2 identical detectors on the the same FW.

A test bed across the board, imho, is somewhat an unrealistic expectation. We need more people that are willing to invest a little time and resource to do their own testing following some simple set ups as suggested above.

Try some back roads and set up some realistic situations that any Leo would use against you. Mounting a radar gun on top of an overpass and shooting straight down a highway is kind of cool, but not something that’s always common. Not here anyways where detectors are legal.

Take multiple runs and post the results. The more the better.


Sent from my iPhone using Tapatalk
 

Deacon

TXCTG
Premium Member
Advanced User
Intermediate User
Newly Registered
Nov 13, 2016
7,427
91
Hill Country, TX
#7
Seriously, it’s all something I’ve thought about. The ideal, short of a custom built facility on a large plot of land, would be essentially an old air field. Plenty of length for different angles of attack and such, no traffic, 100% consistency of placement and reasonable consistency of conditions, etc. The only thing better with be a large race circuit with significant elevation changes, like Sochi or even Austin.
 

Kennyc56

Ford Freak
Premium Member
Advanced User
Intermediate User
Beginner User
General User
Newly Registered
Feb 22, 2018
1,323
21
56
Eastern NC
#8
I look at it like this, when your out in the real world getting popped with I/O, whoever is on the other end of the gun doesn't sit around and wait for perfect conditions! That's what bench testing is for! All you can do is find a way to make each run as close to the last one as you can. After you run the same detector on the same course a few times, you'll get a pretty good feel for when something goes wrong during the run. That's why we call them real world tests!
 

CJR238

-CMS Article Admin-
Administrator
Premium Member
Advanced User
Intermediate User
Beginner User
General User
Newly Registered
Oct 30, 2010
26,219
27
CT. MZDA 3. ALP, V1/RedLine/Waze
#9
Seriously, it’s all something I’ve thought about. The ideal, short of a custom built facility on a large plot of land, would be essentially an old air field. Plenty of length for different angles of attack and such, no traffic, 100% consistency of placement and reasonable consistency of conditions, etc. The only thing better with be a large race circuit with significant elevation changes, like Sochi or even Austin.
Airport would not work. Too open and would need to calculate the curve of the earth with range on an open field. But if we could put in fake houses, trees, foliage, guardrails maybe.
 

Deacon

TXCTG
Premium Member
Advanced User
Intermediate User
Newly Registered
Nov 13, 2016
7,427
91
Hill Country, TX
#10
Not really. Yes, it’s open, but you’re not testing exclusively to find a maximum possible straight-line range (e.g. 13 miles in Oklahoma somewhere). You’re comparing detectors to each other, with buildings to hide behind and taxiways to set up curves, etc. It’s a controlled environment with no unpredictable traffic, nosy neighbors, or other such distractions, and it’s a static environment you can set up identically with a high degree of consistency over time.
 

CJR238

-CMS Article Admin-
Administrator
Premium Member
Advanced User
Intermediate User
Beginner User
General User
Newly Registered
Oct 30, 2010
26,219
27
CT. MZDA 3. ALP, V1/RedLine/Waze
#11
Not really. Yes, it’s open, but you’re not testing exclusively to find a maximum possible straight-line range (e.g. 13 miles in Oklahoma somewhere). You’re comparing detectors to each other, with buildings to hide behind and taxiways to set up curves, etc. It’s a controlled environment with no unpredictable traffic, nosy neighbors, or other such distractions, and it’s a static environment you can set up identically with a high degree of consistency over time.
Agreed.

But you would have to build more buildings, create curves with obstructions, add trees and foliage, hills, overpasses. Sounds like a lot when you can buy/rent a more realistic place with all the things needed.

One of the best places we tested was in a closed corporate park one member had access to. Was a great day of testing: https://www.rdforum.org/index.php?threads/7145/
 

Bulletbling

PSL +5
Intermediate User
Beginner User
General User
Newly Registered
Oct 25, 2015
358
0
#12
Probably the best reason a testbed scenario is unrealistic. You can throw temperature in there as well.

Even the placement of a radar emitting devise. Such as inside a warm vehicle, or an antenna mounted outside changes some things.

The best is constant testing in different environments and taking an average over multiple runs as we do now.

We have wild DA’s (Density of the Air or atmosphere) where I am. If you race you know what I’m talking about.

I can go out Friday and get a DA of 2200 ft. Which is correct for my location. I’ll go out Monday and the DA will be equal to 4400 ft above sea level.

I have to use the Dragtime app to bring the results down to sea level to see if I made any gains.

If we see results in about’s 300 ft of each other over various tests across the continent, i would consider them a good representation of what each detector is capable of.

The other issue is the capability of each detector. We do know that there are the occasional hot copies that comes out and squew’s the results between 2 identical detectors on the the same FW.

A test bed across the board, imho, is somewhat an unrealistic expectation. We need more people that are willing to invest a little time and resource to do their own testing following some simple set ups as suggested above.

Try some back roads and set up some realistic situations that any Leo would use against you. Mounting a radar gun on top of an overpass and shooting straight down a highway is kind of cool, but not something that’s always common. Not here anyways where detectors are legal.

Take multiple runs and post the results. The more the better.


Sent from my iPhone using Tapatalk
You mean density altitude (pressure altitude corrected for nonstandard temperature) I think, right?

If so, I get what you mean. It can vary wildly and that could affect results.