originalfoo / sensorarray

Open source sensor array

Project Goals

Develop a low cost (less than $500) wide-spectrum environment sensor with the following capabilities:

• Light - how sunny is it?
  • IR, visible, UV?
  • intensity?
  • how much is reflected from the ground?
• Clouds - is it cloudy?
  • what sort of clouds?
  • cloud base height?
  • velocity of clouds?
  • affect on solar and wind power?
• Wind - in immediate vicinity of sensor
  • direction and speed?
  • wind chill factor?
• Gas - what gasses are in the air?
  • determine the type and volume of each gas, both in immediate vicinity and also in upper atmosphere
• Moisture - how much moisture in the air?
  • immediate vicinity and also upper atmosphere
  • is it foggy/misty?
  • what is the dew point?
• Precipitation - rain, sleet, snow, hail
  • classify hydrometeor type, size, velocity and distribution
• Lightning - is there an electric storm?
  • storm size and severity
  • storm direction and distance
  • ideally determine if it's IC or CC
• Fire - are there any fires in visual range of sensor?
  • optical sensor to detect flames for early visual warning
  • gas and/or particulate sensor would detect smoke
• Particulate - what size of particles are floating in the air?
  • PM2.5, PM10
  • ideally classify common particulate (eg. pollen)
• Temperature - how hot or cold is it?
  • how much heat from the sun?
  • how much heat from the ground?
• Pressure - what is the air pressure?
  • determine height above sea level
• Earthquakes - is the ground shaking?
  • 3-axis seizmometer
• Sound - how noisy is it?
  • which direction?
  • which frequency bands?
  • how much subsonic and ultrasonic noise is there?
• Radiation - safe or dangerous?
  • alpha, beta and gamma ionising radiation?
  • x-rays?
  • muons? (cosmic radiation)
• Location - where is the sensor?
  • geo co-ordinates?
  • altitude?
  • orientation?
• to be continued

sensorarray's Issues

Heliotrope actuator research

FPV tilt camera Quad UK

Rubber seal below photonics dome

The rubber sealant ring between the photonics dome and photonics base must be UV resistant, mould resistant, etc. Important that it doesn't degrade at all as it will be the 'weakest link' in hermetically sealing the photonics module.

Dynamic calculation of time boundaries for bright sunshine checks

It should be possible to determine what portion of the day bright sunshine is possible within. At the very least, constrain between dawn and dusk. This will reduce number of sensor readings, and thus energy consumption, and also facilitate earlier reporting of daily averages.

Wind turbulence around the sensor

This sort of thing could cause big problems with the anemometer:

FPGA

Looks like at least one FPGA will be needed to do DSP on hypercubes from the hyperspectral cam in photonics module.

Key requirements:

• Lifespan of erase/write/read cycles
• Plenty of I/O
• Environment resilience (particularly temperature)

Heat management

Body of main sensor array is bit like a car – on a hot day it could be up to 50ºF hotter than the outside air, which obviously causes problems for temperature readings, etc.

Current plan is to have a double Stevenson Screen, made from white or other light-coloured plastic to reflect some heat (can't be shiny as that would start fires) and act as an external "chimney" to updraft excess heat and keep it away from body of sensor array.

Potential issues:

• How/where to take anemometer and dry bulb temp readings?
• Heat column rising up at sides of photonics array could fubar thermopile readings

GPS must provide 1PPS

A GPS with a 1PPS (one pulse per second) output is desirable as it allows much more accurate time keeping.

Note: Don’t use modules for GLONASS or Compass as they have other time bases. The European Galileo might work in the future, as it’s clocks are synchronized with GPS

Nephelometer sensors

Sharp GP2Y1030AU0F

Notes:

1. Still under development, specifications liable to change
2. GP2Y1030AU0F is the only Sharp dust sensor that can differentiate between PM2.5 and PM10

• Min particle size: ? μm
• Output: UART - can discriminate between PM2.5 and PM10 👍
• Supply voltage : DC 3–5.5V 👍
• Power consumption : 35mA 👍 👍 👍 👍
• Operating temperature range : -10 to +65°C
• Operating humidity range : ?
• Warm-up time: 5 seconds 👍
• Lifespan: ~5 years (due to optics degradation) 👎
• Approx cost:

Oktameter research

Find a simple, low-power way to determine cloud cover - ideally in Oktas - as this will enable more costly cloud photogrammetry readings to be skipped in cases where there is no cloud or where the sky cannot be seen (eg. due to heavy fog, snow, etc).

Cloud presence can be determined at night using a Seebeck TEG sandwiched between two metal plates. If sky is clear, the top plate will be cooler than the bottom plate. This could double up as power generation during daytime?

A thermopile sensor filtered to wavelengths above 5µm (ideally 8–14µm). Other methods for measuring solar radiation might be valid.

Atmospheric water vapour can affect readings in the 5–8µm band but overall has little effect on sensing quality (source).

Trigger fire warning based on heat?

If ambient temperature rises above 80ºC, is it safe to consider there is a fire or other unexpected heat source nearby? This would enable fire detection even in cases where other sensors (optics, gas, particulate) are malfunctioning.

UEXT port

If there are any spare pins on the main processor, it would be nice to have a UEXT port to allow easy modding/extensibility.

Integrate with WeeWX

http://www.weewx.com/

It's an open source weather station app in python that displays graphs, uploads to popular weather sites, etc.

False-positive optical fire detection

Optical fire sensor test rig was triggering without presence of a fire. Sun was low in sky, shining over fence through frosted window.

How to prevent this type of scenario? How to detect false positives?

Make circular baseplate completely detatchable

When using the array in drone or handheld mode, it would be useful if entire baseplate could be detached to reduce weight and bulkiness.

Baseplate will primarily be used for:

• Anemometer?
• Leaf wetness?
• Neopixel ring
• Oktameter (peltier-seebeck)
• Scintillometer?
• Piezo harvesters
• Connection to mounting pole

Magnetic interference on magnetometer

NFC for casing identification

Casing could contain an RFID tag defining what sort of casing it is?

• No casing -> assume handheld?
• Normal casing -> assume wilderness/urban
• Advanced casing -> assume ship
• Drone casing -> drone
• ...

Will need to have way of interfacing with custom casing (eg. someone designs their own casing for some reason).

Some casing will have effects on sensors, eg. acoustic.

Optical fire sensor research

• How much energy generated when no fire?
• How much energy generated when there is a fire?
• When no fire, can energy be harvested?
• Who makes it?
• Data sheet?

Disable optical fire sensor if particulate/gas sensor detects fire

If particulate/gas sensors are detecting fire, the optical sensor should be disabled to conserve energy (or its output redirected to energy harvesting).

Lightning triangulation

The lightning sensor determines approximate range, but not direction.

To determine direction, triangulation will likely be required with the aid of additional lightning sensors. This will require two or more sensors to be within calculable on-scale range of a lightning strike and also track the time of the strike to a sufficient degree of accuracy (as there can be multiple strikes within the space of a minute).

The speed at which EM interference (used by the sensor to detect lightning) travels through air will be affected by things such as atmospheric moisture, temperature, terrain, etc., with will add further complexity to the calculations.

Hygrometer research

Some useful general infos

Sunrise/sunset distortions of ground-facing photometers

Make sure any ground-facing photometers aren't affected by lateral light from sunrise/sunset. They should only measure light reflected or radiated by ground.

Ionising Radiation MEMS sensor - vendor change

Original vendor sold the product to an as-yet unknown IC manufacturer:

DearMr. Fraser

We wish to inform you that Teviso Ltd. has sold their radiation sensor product line. The new owner is a well established company in the field of sensors and their goal is to market the Teviso sensors under a new label soon. Based on the stipulated non disclosure agreement we do not provide any information on the new owner.

Our thanks go to our customers and business partners for your confidence in Teviso Ltd. and their products.

Best Regards
Roland Glaser

Need to try and find out who new owner is and assess whether pricing / specifications have changed.

Pyranometer thermopile under optic dome

Although entire photonics module is under a dome, better average readings can be obtained by having the thermopile under its own dome within the main dome. In particular, this will prevent small drafts caused by heliotrope movement from distorting the thermopile.

Device orientation

The device needs to know which way (magnetic) North is - is a compass in an IMU sufficient for this task or should installation instructions provide guidance of how to orient the device?

SDK API requirements

Just a dumping ground for things that are going to be required in the software libraries (using JS notation for clarity)...

• .hasDayNight() - is there an active day/night cycle in the current location at the current time
• twilight methods: next twilight, duration, different stages, etc
• some way to measure/determine expected power consumption for each sensor reading

Dome for photonics module requirements

Specifications

Note: Incomplete and likely to change

• Wavelengths: IR, Vis, UV -- depends what optics are in the dome
  • Will need to know which wavelengths are blocked to assess if they will cause problems with any of the photonics
• Internal coating: Preferably anti-reflective (≤2% reflection)
• External coating: Ideally hydrophobic
  • Need to check optical transparency properties of hydrophobic coatings
  • Need to check that hydrophobic coating won't bork the rainlight sensor
• Transmittance: ≥ 98%
• Diameter tolerance: ±0.1mm or better
• Thickness tolerance: ±0.2mm or better
• Weight: As light as possible (drone mode)

Power management research

Things of interest:

• TL431ACLPG shunt regulator diode - given a Vref and a V, if V <= Vref then Vout = V.
• LMS33460 - similar sort of thing to above, but V is fixed to 3V; also has built-in hysteresis to prevent it flapping due to noise.

Create rough sketch of sensor locations

It's likely the sensor array will be compartmentalised; need to work out rough grouping of sensor locations in the array. Key consideration at this stage is effects of ambient and local environment (eg. #10).

Bhangmeter research

A bhangmeter is a non-imaging radiometer installed on reconnaissance and navigation satellites to detect atmospheric nuclear detonations and determine the yield of the nuclear weapon.[1] They are also installed on some armored fighting vehicles, in particular NBC reconnaissance vehicles, in order to help detect, localise and analyse tactical nuclear detonations. They are often used alongside pressure & sound sensors in this role in addition to standard radiation sensors. Some nuclear bunkers and military facilities may also be equipped with such sensors alongside seismic event detectors. – source

As the sensor will have a bolometer, magnetometer, infrasound barometer, and other sensors, might as well enable it to detect nukes too :)

Showreel pondering

If product is launched on kickstarter/indigogo, will need a kickass intro video to illustrate some of the vast range of features. Almost certainly going to need 3D rendering in there.

Sini (just watch this video) looks pretty awesome - maybe they'd be willing to do something if we provide them marketing in return?

Allow external power via USB

Similar to how many arduino boards work; USB can be used purely to provide power if desired.

Will facilitate increased level of active measurements and telemetry; useful for when the sensor array is used as central hub with lots of other sensors connected to it.

Will ideally need some way to know if the power is permanent, limited or transient?

Integrate with PWSWeather

https://www.pwsweather.com/

Ideal Anemometer placement would conflict with photonics module

Placement:

• Can't place anemometer at top, because it would get in way of photonics.
• Can't really place it at bottom, because it would be affected by mounting rod
• Can't really place it in middle as it would get in the way of wires between modules, and weaken design, not to mention issues relating to Stevenson Screen.

Tech:

• Ultrasonic is do-able, but not sure how to process the results
• Microwave might be interesting to investigate - it's not sound, so unaffected by wind?
• How does heat affect things ultrasonic/microwave?
• Could use heated wire approach? - eats too much power
• Normal spinning cup things will freeze in winter

Work out how to test lightning detector

In order to fully utilise the lightning detector, I need controlled testing environment to test triangulation, etc.

How the heck do I create controlled lightning?

Integrate with Weather Underground

https://www.wunderground.com/

Integrate with WeatherCloud

https://weathercloud.net/

Tide clock

Determine low/high tide (times, and current) based on position of moon.

Useful for coastal locations, and possibly also ships?

WMO reporting precision and intervals

source - p35

Infrasound sensor

Cheap MEMS barometers (like the barometer that is already planned for use in the sensor array) can be used to measure infrasound; from 0 Hz up to about 76 Hz. source

Infrasound (<20 Hz) from above-cloud discharges has been detected on Earth (Farges et al, 2005), which suggests a new technique for planetary TLE detection. – source

So, potentially, an infrasonic fulmenometer.

Infrasound can also be used to detect earthquakes (basically what a geophone does).

Natural events: infrasonic sound sometimes results naturally from severe weather, surf,[6] lee waves, avalanches, earthquakes, volcanoes, bolides,[7] waterfalls, calving of icebergs, aurorae, meteors, lightning and upper-atmospheric lightning.[8] Nonlinear ocean wave interactions in ocean storms produce pervasive infrasound vibrations around 0.2 Hz, known as microbaroms.[9] According to the Infrasonics Program at NOAA, infrasonic arrays can be used to locate avalanches in the Rocky Mountains, and to detect tornadoes on the high plains several minutes before they touch down.[10] – source

Dome for disdrometer RFIR?

Assuming that a rear-facing IR TOF sensor (and corresponding IR LEDs) will be used for the disdrometer, it would make sense to protect it from elements with an IR-transparent cover, something like this (obviously smaller):

Find way to implement a scintillometer

Normally they need two devices a mile or more apart, but it should be possible to reinvent the wheel and come up with something significantly simpler and cheaper.

Optical lightning sensor research

Wavelength to check? ~777nm source (related)

Ball lightning can be detected via same frequency – source

Mounting pole

Need to determine how to mount the sensor array, particularly as the pole will connect to the ground facing module at the base of the array and thus potentially affect sensors in that area.

Common poles seem to be about 3.3cm or 1.25" outer diameter. Weather stations are usually affixed to the main pole via a smaller pole.

Blitzortung / LightningMaps

Global realtime network of lightning strikes: http://en.blitzortung.org/

SMD solar cells

Despite worst brochure ever, the IXYS IXOLAR SolarBIT cells look ideal - and are the only ones I've found that are SMD.

Will need at least 4 of them (maybe 6?) on top of heliotropic platter, and ideally 2 underneath.

Each cell on top of the platter will reduce heat reflectance, resulting in increased heat in the photonics module.

For most of the day the camera will be facing away from sun and tilted downwards, ensuring that the top of the platter is facing the sun.

Hyperspectral camera lens, iris, etc...

Mount:

• Likely use an S-Mount (commonly used for CCTV) as it's inexpensive compared to C-Mount and plenty of stuff on the market
• However doesn't look like there are any options for autofocus etc?

Cool stuff:

• Chemical iris, and another one
• bunch of other cool stuff - look in phase 2 as well

Hyperspectral cameras

Most of the sensors required for the heliotropic platter could be condensed in to a single hyperspectral camera - if suitable components can be found.

Sensors

There are two main options identified so far (from cameras list):

• iMEC– ~600-1000 nm - already in production and purported to be affordable (although I've yet to see a price list). Existing cameras based on the tech are available, including very small ones that would fit within the tight confines of the photonics module.
• Hamamatsu – 165 to 1100 nm - easily the best option at the moment but poses many technical challenges (particularly temperature control).

Possible future contenders (not yet in production as far as I can tell) - all of these are targeting mobile phone market; tiny, inexpensive, low power and easy-to-integrate sensors:

• VTT
• PARC
• Unispectral - note: Apartheid regime

More basic options including limiting imaging to multispectral, just IR or very low resolution (often single pixel):

• Melexis - FIR sensor which works within desirable specifications (especially temperature).
• ULIS - thermal imaging cameras, much better resolution than Melexis
• Espros - like an RGB sensor, but does 400 to 900 nm (in currently released product as far as I can tell; earlier press releases stated there would be two variants: 387 – 903 nm, and 776 – 1064 nm)
• AS7262 and AS7263
• Pyreos - very low power consumption
• GridEYE – 8x8 thermopile array

Tech issues

Many of the existing hyperspectral cameras have very constrained operational temperature specifications which would make them unsuitable for use in weather extremes.

They also consume lots of energy, in the order of hundreds of milliamps. This severely limits the number of images that can be generated per day due to energy harvesting constraints. We would likely need a supplementary external power source (larger PV array for example) to operate at full capability.

Processing raw data would likely require either a FPGA for realtime processing, or remote server for post-processing.

Transmitting hypercube files would eat up bandwidth and power, especially over cellular connections.

Ideally we'd want to merge spectra in realtime; similar to infragrams, but with custom spectra being applied to RGB image. This would greatly reduce telemetry bandwidth and provide images in a simple format.

In cases where we do want to retain hypercube we would likely need either additional RAM for buffering or limit the number of spectral layers per capture.

Real time camera mode (eg. for drone use) would likely prove troublesome.

Hyperspectral camera filters?

Polarised filter lens

A polarised filter would help eradicate reflections.

If it could be motorised (ie. to turn it), it would facilitate a Polarimeter - particularly useful for hand-held or lab use.

Optical filter/lens wheel

A rotary series of filters/lenses - a bit like a holga lens; one lens selected at all times:

• no lens – for unadulterated images
• Long exposure with strong UV filter – good for scenic shots or and to prevent over-saturation
• Macro with UVB filter - particularly useful for optical inspection of outer rim in photonics module
• Wide-angle - for landscapes
• Fisheye - for cloud photogrammetry

If room:

• Tiltshift with strong UV filter - for fun

PCB RF shielding fence and covers

RF shielding fences might be required around thermopile sensors for the oktameter.

Less particle shielding might be better

Due to Bragg Peak it might be better to have less shielding from ionising radiation than more - particles will do less damage if they just pass straight through the sensor array compared to those that are stopped within it.