A malaise trap is a tent-like structure that is designed to capture and preserving flying insects.
Most malaise traps are made of PET (Polyethylene terephthalate) netting which insects will fly into and are then redirected by wind into a 500mL HDPE bottle containing Ethanol as a preservative.
The collected insect specimens are then shipped to the University of Guelph's Centre for Biodiversity Genomics where PCR DNA sequencing takes place. The use of high-throughput PCR (Polymerase Chain Reaction) is performed on tissue samples from the insect specimens and the DNA barcoded by CBG's Genomics Unit.
The PCR sequencing that takes place involves the extraction of a DNA barcode. The exact gene region used for the differentiation between insect species is mitochondrial DNA (DNA inherited from the mother) in the cytochrome c oxidase subunit I region (COI).
After sequencing, the collected specimens are then assigned a BIN (Barcode Index Number) based on the sequenced barcode region.
Sampling with malaise traps in the Arctic generally occurs in the late spring when insects begin emerging and continues throughout the summer and fall.
This means that most malaise traps and by extension, most of the collected DNA barcodes from those traps are limited to areas that can be regularly accessed by vehicle in most areas of Nunavut.
As a result, most of the collected specimens are taken within a limited geographical area near communities.
From the perspective of detecting invasive species, this makes it more difficult as species that arrive by sea lift or by any other human means permeate an area around those communities.
This isn't demonstrated yet, but based on initial analysis of DNA barcodes collected in the Kitikmeot region, that does appear to be the case.
That is, unless we have insect species and communities within those species that exist everywhere in Canada. A good example is the common house fly, or even fruit flies. Those two examples of synanthropic species are presumed to exist anywhere that there is a human population connected to the rest of the world. The most notable example being fruit flies which are spread by both fruit and produce. However, that does not necessarily mean that those species can exist in the native environment.
The intent of the Arctic Malaise Trap project is to address this gap in knowledge and to collect species across a larger regional areea of the Arctic. To accomplish this, rather than deploying the malaise traps in the late spring, traps will be deployed in the winter instead. To make this possible, the use of low power electronics and solar power are intended to initiate the start of the malaise trap sampling to occur later on in the summer completely unattended.
Insect specimens collected using malaise traps are preserved in Ethanol, and since DNA degredation is tied to ambient temperatures above 5°C, the use of refrigeration is used to preserve that DNA long enough for successful DNA sequencing from those collected specimens to take place.
Coincidentally, in tundra conditions in the Canadian Arctic, both permafrost and the active layer of permafrost are rarely warmer than 4°C. This means that the permafrost is a potential source of naturally occurring refrigeration in the field.
In other words, collected bottles of insect specimens and Ethanol can be lowered deep into the permafrost for long term preservation of DNA.
Most transporation issues in the Arctic occur around the shoulder seasons when both sea and lake ice present barriers to moving people and equipment.
The use of a winter time deployment of malaise traps and the subsequent retrieval of those samples in the winter allow for the deployment of new trap sites when transportation conditions in the Arctic is at the easiest.
Travel by snowmobile in the winter can occur quite easily even to extremely distant locations from communities provided that adequate shelter, fuel, and supplies are accounted for.
In addition, the fact that it is generally well below -20°C in the winter in the Arctic means that the preservation of permafrost core samples is also relatively simple compared to keeping ice cores frozen in summer time conditions. Since a hole will need to be drilled, taking advantage of that means you can accomplish more science per deployment. A single site can house a malaise trap, a recoverd permafrost core sample, and deployed temperature and moisture sensors on a pipe installed in the permafrost.
A permafrost core sampling drill is used to drill a 5-6 foot deep hole into the permafrost through it's active layer (layer of soil which melts in the summer and then re-freezes again in the winter).
A 10 foot ABS pipe is then lowered into the vacant hole with the permafrost core recovered and packaged for further analysis. Rather than simply disposing of the permafrost core sample from the hole, a permafrost core sample can be recovered at the same time. Since the deployment occurs during winter, the natural preservation of the permafrost core sample allows for the sample to be packaged and retrived when ambient temperatures are well below freezing. No field refrigeration or coolers are needed.
An arctic malaise trap is then installed and connected to the top of exposed ABS pipe. The arctic malaise trap itself is a simple 3D printed assembly and an adapter connected to the most common commercially available malaise trap.
The critical improvement over the existing malaise trap design involves the use of low power electronics to serve as a time delay for the opening of a valve mechanism on the malaise trap and to also lower the collected specimens in a 500mL HDPE bottle to the bottom of the ABS pipe where it can sit in the permafrost for long term preservation until it's eventually retrieved.
Since the trap is deployed in the winter, this alleviates a major issue with reaching distant and remote sites in Nunavut. Snowmobile and travel by kamotik/sled is the general method of travel in Nunavut in the winter and is not only well understood, but well practiced in nearly every community by Inuit.
All that is then needed is for the malaise trap to trigger the collection of specimens in the late spring at a pre-determined time, seal a 500mL HDPE bottle, and lower the collected specimens into the permafrost.
The low power electronics on-board the Arctic Malaise Trap will utilize onboard WiFi capabilities and a deep sleep/wake function of the ESP8266 microcontroller to wait for a specified interval before connecting to a wireless network served by a base station running a network time protocol (NTP) server. Once the specified trigger day and time is reached, the Arctic Malaise Trap will open a valve for a 500mL HDPE bottle of Ethanol. In the 24/7 daylight in the Arctic, and depending on ambient temperature and humidity the Ethanol will evaporate at a rate of 20mL per day. That gives each individual Arctic Malaise Trap a sampling interval of approximately 25 days. Once those 25 days have concluded, the 500mL HDPE bottle of Ethanol is securely closed and lowered into the permafrost for long term preservation.
Rather than deploying just one Arctic Malaise Trap per site, the use of multiple traps in a single location would allow for the coverage of an entire sampling season. Each trap would have a set point to open the valve, and to close the valve.
At the conclusion of the sampling season, all of the 500mL HDPE bottles should have been lowered safely into the permafrost for the long term preservation of insect specimens for later retrieval.
Since DNA is preserved in Ethanol quite well for durations of 2-3 years at 4°C, this means that each site can be serviced and the specimens retrieved when it's convenient.
The main compponent is the malaise trap itself. This is a tent-like structure that measures approximately 6 feet long by 3 feet wide by 6 feet tall.
The tent itself funnels flying insects into the very top of the malaise trap which leads to an inverted 500mL HDPE bottle that has been modified to include a hole through the PET fabric of the trap leading into the inverted bottle itself. This 500mL HDPE bottle has a coupler which securely attaches the inverted bottle and another 500mL HDPE bottle together.
The use of low cost, low power ESP8266 WiFi enabled microcontrollers will be used to trigger the opening, closing, and lowering of the 500mL HDPE bottle of Ethanol into the permafrost for preservation.
The malaise trap itself and the coupler will be used to interface with an adapter sitting atop a 10 foot long 3" ABS pipe. Within that 3D printed adapter, there will need to be the low power electronics to open, close, and lower the 500mL HDPE bottle into the permafrost. Additionally, the 3D printed assembly will also need to house the batteries and a small 6-12 Volt solar cell.
While the exact mechanism for opening and closing the bottle is under development, several prototypes to evaluate the stepper motors performance to determine the most reliable method of performing these two tasks will need to be conducted.
In addition, stepper motors, magnets and the use of sealed magnetically activated Reed switches will be used to determine the position of the malaise bottle within the 3D printed ABS pipe adapter assembly and 3" ABS pipe.
A central base station with a connected Raspberry Pi will also need to be deployed to serve a limited range of deployed traps. The Raspberry Pi will operate continuously and supply the deployed traps with the corrected date and time via a connected GPS module.
Within line of sight of the base station, the extremely low power ESP8266 WiFi microcontrollers will connect to a Raspberry Pi base station to check the current date and time. Once a specific date is reached, the Arctic Malaise Trap will open a valve in the 3D printed adapter for the connected 500mL HDPE bottle containing 500mL of Ethanol. Once a specific sampling period has occurred, the valve will then close, and the sample then lowered into the permafrost for the long term preservation of the insect specimens and their DNA.
The Wemos D1 Mini is an ESP8266 WiFi microcontroller that has a deep sleep function which uses almost no power. Of the available ESP8266 options, the ESP8266 Wemos D1 Mini uses the least amount of power in deep sleep mode. When in deep sleep mode, the ESP8266 microcontroller sets a watchdog timer that triggers the re-start of the ESP8266 to occur after a specified interval of time. The exact amount of power consumed is 20µA at 3.3VDC, or 66µW.
However, the maximum time that the deep sleep function can be used is 71 minutes, or 4,294,967,295 microseconds as determined by the 32-bit register used by the watchdog timer that is used to count the time.
Because of this, the ESP8266 can instead be set to wake precisely every hour to check the current date/time via the WiFi connection to the Raspberry Pi BaseStation. During that operation, the ESP8266 will use around 70mA when idle and as much as 400mA while under load. However, the ESP8266 will not be under extensive load and the idle power used is not far off the 70mA measurement.
In order to power the ESP8266, the microcontroller can be powered either by a single 18650 Lithium battery, or off several AA Ultimate Lithium batteries.
The Energizer AA Ultimate Lithium batteries are much better suited for long term operation in cold temperatures, however they are not rechargable. This would mean the AA batteries for the ESP8266 microcontrollers would need to be replaced once per season.
However provided an 18650 lithium battery cell could be insulated, heated, and recharged via solar it may potentially be the better long term option. Both options will need to be tested and evaluated.
First, let's establish some numbers.
There are 24 hours in a day and 60 minutes in an hour. So, 24 hours * 60 minutes per hour = 1440 minutes per day
Next, what are amps, milliamps, and microamps? They are all measurements of Current.
As per Ohms law, Power is measured in Watts and is equal to Voltage multipled by Current.
For example: 12 VDC * 1A = 12 Watts. If you're familiar with USB devices, they generally run at 5VDC and 500mA. So, 5 VDC * 0.5 = 2.5 Watts.
If you're familiar with both USB-C and modern smart phones then you've likely seen USB-C chargers advertised as capable of fast charging. That is accomplished by either increasing voltage or current. To maintain backwards compatibility with older USB devices, the voltage is not increased from 5VDC. Rather than increasing voltage, USB-C accomplishes the delivery of more power by adding supplied current in the form of additional wires to a USB cable and USB-C or USB interface connector.
So, 5VDC * 1A = 5 Watts. 5VDC * 2.5A = 12.5 Watts. And so on.
This is an important consideration for the purpose of the long term operation of the ESP8266 WiFi Microcontroller because while it does operate on an extremely low amount of power, the total amount of power needed for it to run as measured in Watts is not zero.
Ampere: 1 Ampere = 1000mA
Milliampere: 1mA = 0.001A
Microampere: 1µA = 0.001mA
Idle, WiFi on: 70mA @ 3.3VDC, or 0.07A. 0.07A * 3.3VDC = 0.231 Watts.
Under load, WiFi on: 400mA @ 3.3VDC, or 1.32A. 1.32A * 3.3VDC = 4.356 Watts.
Deep Sleep: 20µA @ 3.3VDC, or 0.00002A.
0.00002A * 3.3VDC = 0.000066 Watts.
While 0.000066 Watts is basically nothing, accumulated over 365 days that small number does turn into a much larger number.
The ESP8266 WiFi handshaking connection process while connecting to a WiFi network typically takes between 1-2 minutes.
1-2 minutes * 24 hours = 24 to 48 minutes per day spent connecting to a wifi network to check the current time.
The typical power use per day is therefore 24 to 48 minutes at 70mA, or 24 min * 0.07A = 1.68 Watts on the low end.
And 48 * 0.07A = 3.36 Watts on the high end. 24 minutes to 48 minutes are spent connecting to the Raspberry Pi BaseStation.
However, we also still need to include the deep sleep power consumption before coming to the final amount of estimated power needed to run the ESP8266 for an entire year.
1440 minutes - 24 minutes = 1416 minutes are spent at 3.3VDC and drawing 0.00002A of current.
0.00002A * 3.3VDC = 0.000066 Watts
0.000066 Watts * 1416 minutes = 0.93456 Watts per day is spent at low power during deep sleep.
Next, we add the low power deep sleep power use to the idle wifi enabled power use and the number is: 1.68 Watts + 0.93456 Watts = 1.773 Watts per day on the low end.
We also need to calculate for the amount of power used on the high end, 3.36 Watts + 0.93456 Watts = 4.294 Watts per day
It's unknown whether the ESP8266 will draw more power when running the stepper motors, however initial power measurements performed with a USB Watt Meter indicate that this is not the case. The average measured power use during operation of the 28BYJ-48 stepper motor ranged from 72mA to 76mA. Which is not much more than the estimated 70mA.
Therefore for the ESP8266 to run for 365 days, or one year:
1.773 Watts * 365 days = 647.15 Watt hours needed
4.294 Watts * 365 days = 1567.31 Watt hours needed
These numbers are important when compared against the two potential long term battery sources in the next section. Batteries are generally measured in total deliverable Amp hours, or Watt hours.
An 18650 Lithium polymer battery has a nominal voltage of 3.7VDC and typically has around 2200mAh of available power. That's 8,140 Watt hours of power available.
The Energizer AA Ultimate Lithium battery has a nominal voltage of 1.5VDC and typically has a capacity between 4,448 Watt hours to 2769 Watt hours depending on the load in Amperes ranging from 0.1A to 3A respectively. And because it'd be setup in either a 4 cell arrangement at 1.5VDC rather than 3.7VDC, it is a slightly different configuration than running the ESP8266 off a single 3.7VDC 18650 cell.
While the 18650 Lithium polymer battery has more overall capacity and is also rechargable, it also has much less desirable discharge characteristics in colder temperatures.
For example, an 18650 Lithium Polymer battery cannot effectively charge at temperatures below -15°C. In addition, drawing power in cold temperatures reduces both it's overall available capacity and the total lifespan of the battery over time. The typical reduction in total available capacity after prolonged use after a long period of time is around 60% of it's overall capacity.
2200mAh * 0.6 = 1320 mAh @ 3.7 VDC, or 4884 Watt hours of available power.
Luckily, that is still within the margin of ~1600 Watt hours needed to operate the ESP8266 during both it's deep sleep cycle and during WiFi connected operation. However, cold weather testing would also need to be performed to determine it's discharge curve in temperatures below -15°C since the 18650 cell will spend most of it's time in temperatures below -30°C.
Therefore, it may be worthwhile to test two configurations for powering the ESP8266 microcontrollers.
The AA configuration involves the use of a 4 cell battery holder and is relatively straight forward. Simply connect the positive and negative leads from the battery holder to the ESP8266 5VDC power input. The Wemos D1 Mini ESP8266 WiFi microcontroller has an input voltage between 4VDC and 6VDC and four Energizer AA Ultimate Lithium batteries rest within that input range.
The rechargeability of the 18650 Lithium Polymer cell would involve the use of a TP4056 integrated circuit that is also connected to a 6 Volt solar cell. The battery would simply need to use the available sunlight to recharge in the summer when temperatures are above freezing, and collect enough of a charge to have enough power to run over the winter.
A great instructables on using solar power and a TP4056 integrated circuit to power an ESP8266 is available here.
There's a handy website illustrating how much available lifespan you can get out of an ESP8266 in this exact sort of configuration with an 18650 cell.
The exact time that it takes for a stepper motor to lower a 500mL bottle filled with Ethanol to the bottom of a 10 foot ABS pipe would need to be performed. How many steps does it take? How much power is used?
In addition to lowering the bottle, the amount of power and the time that it takes for a valve to open/close would need to be performed as well.
This would give you the minimum and maximum amount of power needed to perform those three specific tasks.
Open valve/flap to start collection. Close valve/flap to end collection, and finally lowering the samples into the permafrost.
An alternative approach could use a larger solar panel to power a small supercapacitor bank providing the instaneous voltage and current needed to run the stepper motor in precisely stepped increments. Since supercapacitors, or ultracapacitors utilize carbon electrodes they are considered solid state. No liquid chemicals or electrolyte are used in their construction.
As a result, they have a nominal safe operational temperature all the way down to -40°C.
It would require an in-circuit power measurement to determine when the supercapacitors are in a charged state from the available solar power, and when they are charged advance the stepper motor a specific number of steps until the capacitor bank is discharged, then the microcontroller would simply need to wait and repeat the process. This would alleviate power concerns as the stepper motors do draw a fair amount of power during operation, but it does involve a higher initial cost. However, no batteries would need to be replaced seasonally for the stepper motors, only the four AA batteries for powering the ESP8266 and it's sleep timer.
As per the video, they're extremely low cost stepper motors. Usually available for less than $5 CAD.
With the ULN2003 unipolar stepper driver that comes with them, they're fairly anaemic in terms of torque.
ULN2003: At 5VDC @ 340mA or 1.7W of power used, they can generate about 33 grams of force or 32mNm.
This tutorial covers the 28BYJ-48 stepper motor in detail as well as how to program an Arduino to control them.
The current early stage prototype of the ESP8266 WiFi microcontroller is functional, but the torque is not anywhere near enough to lower a 500mL HDPE bottle of Ethanol into the permafrost. The 28BYJ-48 stepper motors can be modified to run in a bipolar configuration rather than a unipolar configruation using an A4988 or DRV8825 Stepper motor driver.
The 42BYGHW811 NEMA 17 Stepper motor has dimensiosn measuring 42.3mm x 42.3mm x 47.21mm.
There is a 5mm diameter shaft extending 24.5mm from the base. At the base there is a circular ring 2.5mm tall base that serves to fit the stepper motor into a base or mount. It's 22mm in diameter.
The test components for use indoors will be made of PLA filament, however the final components will need to be printed using a better material for use outdoors. This is because they'll be exposed to extreme weather, moisture, and 24 hours of UV light in the form of sunlight in the summer.
A comparison of the various material filament types available is here. Matterhackers article on outdoor filament
ASA is likely the best material, but would require a more complicated printing process.
The controlled dimension of ABS pipe is generally the exterior or outer diameter during manufacturing. The interior or inner diameter is uncontrolled, which means it can vary by a signficiant amount.
As a result, the 3D printed design will need to fit over an ABS pipe in order to utilize the controlled diameter.
After measuring the 500mL HDPE malaise bottle, it's exterior dimensions are 72.5mm. A 3" ABS pipe would likely be too tight a fit for the malaise bottle, a 3D printed holster, and the bead chain used to lower it into the permafrost.
For these reasons, a 4" ABS pipe is the best likely candidate for testing.
In addition, the available 4" ABS pipe comes in 12 foot lengths rather than 10 feet lengths.
Wemos D1 Mini ESP8266: The Wemos D1 Mini ESP8266 WiFi microcontroller is extremely inexpensive. They can be purchased for approximately $5 CAD, or less in bulk.
28BYJ-48 Stepper Motor: The 28BYJ-48 5V stepper motor is also extremely inexpensive, and they are also available for around $5 CAD.
A4988 Stepper Driver: The A4988 bipolar stepper motor driver is also approximately $5 CAD.
NEMA 17 Stepper motor: There are several models of the NEMA 17 available, but generally they retail for between $18 and $28 CAD each.