WWF's Conservation Innovation Awards

Submitted
Gathering votes
Support
Gathering votes
Originality
Refine
Eligible
WINNER

Thermal Imaging to unmask what is in my backyard

To apply thermal imaging and smart software for identification and monitoring of invasive mammal pests (mice, rats, stoats, possums). to assist in achieving a predator free NZ 2050.

What conservation problem are you trying to solve?

To improve the tools for monitoring invasive pest populations by using non-invasive methods. By using low cost thermal imaging sensors combined with software using artificial intelligence (AI) "what is where" pest data can be gathered.Long term in-field monitoring of pest populations is required. By using low power electronics, a long battery life creates an opportunity to monitor pest populations over long periods (weeks/months). Long term monitoring is important for all stages of pest eradication (preliminary stages to determine pest population, during eradication and after completion). Objective data (repeatable, reliable and robust) is collected on animal type/activity and held in a repository. This reduces the reliance on subjective in-field data collection methods. (e.g chew cards need bite marks to be interpreted leading to variable results). By using more data collection devices an improved data pool data can be created. Therefore data on pest populations is more accurate.

How are you going to solve this conservation problem?

To use thermal imaging sensors to gather heat signatures on pest populations and identify with an acceptable level of accuracy the animal type.The thermal image sensor with a field of view (1.5 x 1.5 metres) connected to a micro-processor detects the animal heat energy. By processing the data, near real time information can be provided onAnimal sizeAnimal movement (direction tracking)Type of animal (determined by animal heat signature)When animal was present (time and date ), how long (dwell time) Information of animal heat signatures is required to discriminate animal types. A project at Lincoln with Dr Bruce Warburton will provide this information and will be available for the wider community.A pedestrian tracker was successfully trialled at Otari-Wilton bush that utilised heat signature data on humans. It demonstrated a working concept and can be scaled to predators. It did indicate that an efficient hardware platform and smart software was required for an effective working model.

What makes your idea new and unique?

Utilises non-invasive thermal imaging technology for pest species monitoring and identification. No need for external lighting sources (e.g Infrared light). Thermal imaging sensors are inexpensive and power efficient. (less than $25 and 5mA operating current )Sensor provides enough resolution data for successful object recognition. Low volume sensor data can be processed by a smaller processor.  In-field close to real time identification of target species is possible by using artificial intelligence.Software uses AI that is capable of self-learning. Monitoring pest activity provides a rich learning environment. The software uses existing algorithms and data (species heat signatures) as a foundation and adapts the algorithm depending on environmental conditions.  In-field data collected can be transferred via remote communications using  digital radio technology.  (LoRa). The radio is designed to transfer low volume sensor data and is very efficient over long distances.

Who will use your idea, and how will they benefit?

Any one with an interest in the pest population data. Data from multiple sites held in a central repository can be accessed by groups such as Regional or local councils, DOC, trapping contractors and community groups. Having long term consistent and reliable data that is kept up to date provides users of the data the ability to monitor predator-prone areas more effectively. The design utilises low power technology that can be built into a small foot print (100 x 50 mm). Size is governed by battery. The unit ( Grid-i ) deployed in bush /urban environments can be mass produced at affordable prices. Using light weight rechargeable battery technology (Lithium Ion). Many units can be carried into remote locations and setup as a sensor network connected to a central base station.It can easily be installed by anyone (e.g Predator free community groups, schools) and the effects of trapping on the pest population can be monitored. 

What tasks or activities do you need investment for? How would you spend a $25,000 grant?

Funding will go towards development and testing so that Grid-i can be ready for production. Produce a second prototype run for field trials (5 units)Develop software to record animal thermal signatures and software tools to analyse the data. This work will be carried out with help from Landcare Research in Lincoln.Determine what level of sensor data output allows for reliable target identification.  Expand AI thermal software library for pest targets (mice, rats, stoats, possums)Design and build central repository for holding data. Need outside resources. Setup urban trial in Ngaio, Wellington. This is a community based predator trapping programme (Predator free Ngaio)With past experience working on military products as an embedded software and design engineer and help with extra resources and community support, developing Grid-i into a viable product is achievable. Having another tool in our toolbox will assist in our goal of a predator free NZ.

Are you a New Zealand citizen or resident?

YES

I have read and agree to the Crowdicity Terms of Use, the Conservation Innovation Awards 2017 Supplemental Terms and Conditions, and the Crowdicity Privacy Policy

YES

List five other ideas posted in the challenge that excite you. Why?

1. Cacophony project:

Why: Aiming to catch the final predator for full eradication in our quest for a predator free NZ 2050 will require a full arsenal of tools and new technology such as using sound will become important. By taking collaborative approaches these solutions will become reality.

2. Swimming with E-Coli

 Why: Need better methods to gather information on the health of our waterways. Providing real time information provides valuable information on the current state and trends can be monitored by management to make better informed decisions.

3. Compost and Catch, spreading the free predator movement

 Why: This is an emerging urban trend as more compose bins are present, more predator breeding areas, so anything that keeps the pest numbers down in our environment and encourage the urban population to partake.

4. The Ballen Filter-Preventing Plastic Microfibres from enter the oceans.

Why: Not a widely known topic. Effects of synthetic fibres in ocean environment and food chain is unknown. Stopping entry at the source is a good idea and using a fibre filter is good but needs further investigation. Relying on human intervention to clear fibres would be unreliable.

5. Wall-E 2.0

Robots to do mundane work to clean up the environment is an emerging trend. Applying technology using robotics by the College is great way to engage young minds and learn. The practicalities of designing, building and testing the beach cleaner is the best learning environment.

How could you improve your idea?

By reading the posts there is an opportunity to improve the functionality of grid-i while adding value to a number of ideas presented in the innovation awards.

1) Target unit for domestic environment. Emerging Urban trend (James Prier) more composts providing more opportunity for predators to breed. Idea improved by targeting monitoring small areas of activity around compost heaps, Allow retro-fitting of unit to compost heaps  

2) As a real time information collecter on what pests are where. We can use this information in collaboration with the self-resetting trap "One self-resetting trap that kills both Possums and Rats (Kevin Bain)". The trap is designed to catch large target predators possums, stoats) but open to catching non-targets such as domestic cats, birds. By using  grid-i to correctly identifiy the targets from non-targets it is possible to electonically prime the trap only for wanted targets, and unprimed for non-targets. The idea is improved by  requirement to expand thermal signatures and identification of other warm blooded animals (birds)

3) Terry Moon posted that the movement of pests can be tracked around the home environment with multiple grid-i units. By locating units next to food sources (compost or inside pantry) or points of known traversal. Pest activity can be tracked and the effectivenes of eradication programmes monitored.Idea improved by potential indoor application for device and to expand thermal signatures to domestic pets

edited on Oct 14, 2017 by Gerald Dickinson

Philinda Dickinson Sep 29, 2017

Hi Gerald, An Idea to identify animals using heat is great. How will it work on smaller animal such as mice. You say it worked on humans ok as a pedestrian counter but there is a big difference in size.

Reply

Gerald Dickinson Oct 2, 2017

Hi Linda,
Good question.
The human pedestrian counter tested at Otari-wilton bush could detect objects up to 5 metres away and track their size, direction and speed. Even differentiating the object size (adult vs child) and speed (running or walking) was possible.

For animal tracking it is envisaged to locate the sensor close to a food lure. (Say 1 to 1.5 metres away) so animal is attracted to the area and inside the sensor field of view.. By collaborating with Dr Bruce Warburton at Lincoln we are testing what sensor data is captured for different animal species (possums, rats and stoats). This data will be processed to
1) Determine that different target species have different heat profiles
2) Determine if target species can be remotely identified using thermal imaging.

The results should be available for the wider community to review and use in future thermal imaging projects.

Another goal is to do the image processing for species identification in real time inside the electronic device (grid-i). Information such as species head count along with time/date data will be gathered. This is where the software will use artificial intelligence along with digital signal processing to get the job done.

This process requires a lot of number crunching and needs an efficient powerful processor. This has to be balanced against minimising power consumption for long battery life (More powerful processor leads to more power consumption and leads to shorter battery life) This is where the challenge lies and being worked on at the moment.

Reply

Philinda Dickinson Sep 29, 2017

comment removed

Reply

JessieC Sep 30, 2017

Hi Gerald,
Great idea! How can I share this by email so others get the invitation to register like I did?

Reply

Gerald Dickinson Sep 30, 2017

Hi Jessie,
Easy head to the Crowdicity wesite in link https://wwf-nz.crowdicity.com/
select "WWF Conservation Innovation Awards 2017"
Select Register
Select "Thermal Imaging to unmask what is in my backyard"
Have a read,comment on idea and vote
All feedback would be great

Reply

Jenni Guzman Sep 30, 2017

Hey Gerald,

This sounds really great. I think that something inexpensive and portable, and able to be sized to meet different needs can have a real impact on the community. It's one thing to have a solution, but it has to be sustainable - which this is! Good work.

Jenni

Reply

Gerald Dickinson Sep 30, 2017

Hi Jenni,
Yes always looking at ways in making more sustainable devices. Appreciate your feedback.

Reply

Brent Beaven Sep 30, 2017

Hi. Interesting concept that has a couple of parallel projects in NZ. One group is working on touch screen print recognition, another on using cameras and AI to recognise animals. Would pay to outline the pros and cons of thermal imaging over these other techniques. I would also be interested in hearing how it will deal with the range of temperatures and other climatic conditions that it would be subject to in the field.

Great thinking,

Brent

Reply

Gerald Dickinson Oct 1, 2017

Hi Brent,
Nice to have questions on this topic.
I am in the process of putting together a list of pros and cons to compare the various techniques and what environment issues would impact on TI reliability, Will post this Monday

In the mean time do you have further information on the touch screen print recognition and camera AI projects? Maybe links to projects
Gerald

Reply

Brent Beaven Oct 2, 2017

Hi Gerald. Look forward to seeing the pros and cons.

Print recognition: https://www.lincolnagritech.co.nz/news/lincol...bc-and-radionz/

Camera project: https://wwf-nz.crowdicity.com/post/424822

Brent

Reply

Gerald Dickinson Oct 2, 2017

Hi Brent,
I have updated the document for reviewing the pros and cons of the three technologies. See the attached document in later post.
Gerald

Reply

Gerald Dickinson Oct 6, 2017

Brent
I have added a detailed review on the three technologies, Touch pad (PAWS) optical sensing and thermal sensing. It highlights the pros and cons.
It also highlights mitigating measures that will be used for the grid-i project using thermal sensing.
Makes comprehensive reading.

Reply

View all replies (4)

Bev Clarkson Sep 30, 2017

Great project. We need more efficient and cost-effective monitoring tools like this to target pests in order to achieve a predator-free NZ.
Nifty little gadgets - how are you going to prevent their being stolen?

Reply

Gerald Dickinson Oct 1, 2017

Field equipment being stolen is always an issue. Some ideas would be
1) Physical Restriction such as good old fashioned chain the unit to the tree.
2) Deterrent such as warning label "This device is electronically tracked if moved". This is done by using radio communications. If device is moved the Base station communicating with sensor unit detects a change in signal level. i.e if sensor moved the signal strength level will increase or decrease. Information on disturbance can then be relayed to cell phone to notify user, the sensor unit has moved.
3) Deterrent such as Audible alarm,(e.g. Car/house alarm) In built motion sensor detects movement and sounds alarm.(We can use 2 direction accelerometer sensor) . Have warning label "This device will sound alarm if moved" would act as first stage deterrent.

Reply

(Account removed) Oct 2, 2017

Hi Gerald, this is really innovative yet practical idea. All the best! Cheryl

Reply

Gerald Dickinson Oct 6, 2017

Thanks Cheryl

Reply

Michael Jordanoff Oct 2, 2017

A great idea Gerald

Reply

Gerald Dickinson Oct 2, 2017

Thanks Michael

Reply

Nigel Binks Oct 3, 2017

This sensor could be useful in a range of applications associated with heat signature tracking, certainly benefiting the monitoring of pests and trap efficacy over time.

Reply

Gerald Dickinson Oct 3, 2017

Hi Nigel, Yes there are many future applications for this device once the monitoring and pest identification function are present, new features can be added and controlling of equipment is possible
Examples
1) Trapping of targeted live animals. An animal (target or non-targeted) present inside the cage can be detected and identified where targets are trapped and non-targets only presence information recorded.
2) Automatically dispense aerosol lures into the environment to attract targets to area or deter unwanted targets from area, e.g rat traps setup to capture rats are often made in-effective by mice eating the bait (mice are light therefore undetected by rat trap)

There are many more possibilities.
Gerald

Reply

Andreas Theodosiou Oct 3, 2017





Hi Gerald,
You are detecting heat of the object, what about if the background is the same temperature as the animal temperature ie on a hot day, Would it work very well and how would you differentiate the animal from the background.
kind regards
Andreas

Reply

Gerald Dickinson Oct 6, 2017

Hi Andreas,

Good question, Thermal technology is not effective when differentiating hot objects from a hot background. When of similar temperate it merges with the background. A possible solution is to use another technology such as optical camera and take a dual technology approach. This situation requires the thermal and optical sensors to be located adjacent to each other with both having a similar field of view.
There is a good case study where this approach is taken;
www.mdpi.com/1424-8220/15/5/10580/pdf

It highlights the strengths and weakness of the two technologies when used separately and the benefits when a dual approach is taken. Where one technology has a weakness the other fulls the gap. With good image filtering and merging techniques the combined two resultant images can be merged into one allowing for object extraction.

This issue is less so for invasive pests with nocturnal behaviour. It is likely that animal activity occurs more in the night time where the background temperature is cooler than during the day time.

Reply

Leigh Nicholson Oct 3, 2017

Having been to Zealandia recently and seeing what a pest free NZ is like - I think this idea has great ptential

Reply

Gerald Dickinson Oct 6, 2017

Hi Leigh,
Yes places like Zealandia are national treasures and require continuous effort and resources to maintain the status quo for generations to come. So having long term in-field detection tools to monitor animal breaches within the sanctuary is very important. Having tools that can give near-real time information on pest type and activity is key. This provides the opportunity for timely and efficient resource use for pest destruction. To ensure the right trap is used at the right time.

I appreciate your feedback
Gerald

Reply

John Manning Oct 3, 2017

Gerald
I think the basic idea is good because when eradicating pests its more important to count the live population than to count the number killed. However in order to judge the value of this tool can you say how much its application will cost in comparison with camera systems. Cameras fail because they are expensive and there are few deployed in any study. Will you device improve on that. Secondly when a device is placed the disturbance and scents left behind discourage animals from the area. Can you unit stay out long enough for that problem to be overcome.

Reply

Gerald Dickinson Oct 6, 2017

Hi John,
Great question,So this is a question on comparing the system costs of using thermal and optical monitoring tools for outdoor use (battery operated). Looking at online retailers for optical based cameras (outdoor trail cameras) prices vary from $250 and upwards. These cheaper lower end automatic detection cameras have basic functionality such as object detection, and record for a set duration before returning to a low power sleep mode to preserve battery.
A major disadvantage is wake up time from object detection before recording starts. This can be up to a 1 second and creates a non-capture window, therefore any event in that window is unrecorded. (no good for fast moving objects)
Another disadvantage of trail cameras is the recorded data needs to be extracted and processed (non-real time) usually by human to shift out wanted from unwanted footage or to use a third party processing tool. With large data files this is time consuming and requires a good processing platform with plenty of memory .These optical camera solutions can be expensive and do not provide a full turnkey solution. The trail camera is not targeted towards object extraction and identification.

This is were building an application targeted to identifying invasive mammal pests is the goal of grid-i. By processing images in near real-time it is possible to process the images on the fly and extract wanted object information (such as rat signature detected at this time)
By using a low-resolution thermal sensor taking 10 images per second. (sensor price approx $25) a smaller processor can be used. This resolution is weighed up against ensuring enough information is present in the image for accurate animal detection and identification.The low volume of data captured can be processed by a smaller less powerful processor in near-real time. Having less data puts less demands on the processor and less power consumed hence longer battery life. (Major goal is for long battery life)
A recent trial at Otari-Wilton Bush in Wellington was successful in detecting and identifying basic human objects and movement in near real time by thermal sensing. This provides evidence that capturing and processing low-resolution data images (along with software image processing techniques and Artificial Intelligence) provides a robust solution.

We are looking for grid-i to be price comparable to low end trail cameras and provide a total solution. Having electronics designed to be power efficient to give good battery life yet powerful enough to do animal identification is an achievable goal with technology available today.

Reply

Gerald Dickinson Oct 6, 2017

Hi John,

In answer to your second question when a device is deployed in the field the scent left behind could discourage the animal from the area.This is especially true for scent sensitive animal such as rats. Scents over time will degrade (how long I am not sure of) and be replaced by ambient environmental levels . This is where the device must have a long battery life (such as in grid-i) and remain undisturbed to allow for non-intrusive surveillance.

Reply

View all replies (2)

James Wilson Oct 4, 2017

Any technical development in predator trapping must be supported. It is only by assisting a range of technologies that we will in the end get closer to the ideal. This one looks good.

Reply

Gerald Dickinson Oct 7, 2017

Hi James,

Appreciate your feedback, Yes there will be many tools in our tool box that will help reach the ideal goal.

Gerald

Reply

Murray McLea Oct 5, 2017

Great idea Gerald. The results will be really useful and effective.

Reply

Gerald Dickinson Oct 7, 2017

Murray ,Thanks for your support

Reply

Terry J Moon Oct 6, 2017

Hi Gerald.

I like the potential of your idea. As part of the predator free program in our local area one often wonders how successful it is in my own garden having not caught anything in the trap. Possibly because there is too much other feed around or they are making their way in to the house instead. This would give some sort of "positive" feedback.

Do you see any potential for your idea within the house such as basements, attic spaces and garages as an early warning of any presents of such pests, to confirm a problem exists or even to gauge success of domestic eradication programs within our personal living spaces.

Could it "tune out" normal domestic / human activity to enable this ?

Could the data collected from inside and outside determine if a local problem is in the environment and ventures in to our living space or in our personal living space and ventures out in to the environment ?.


Reply

Gerald Dickinson Oct 7, 2017

Hi Terry,

Yes we have the same problem at home, having an enclosed compost with a warm dry environment with tasty food on tap and a trap located next door with stale peanut butter. I see why the animal chooses the compost over the trap.
This is on my to do list to have a unit monitoring activity around the compost and trap and see what rodent behaviour this uncovers.

Rodent presence detection in spaces in and around the home would be a good application. Great for monitoring the progress of domestic eradication programmes.

Animal identification is based on the thermal signature and the detection application could be tuned to record activity of specific targets (such as what , when and where) and non-targets ignored. i.e. tuned-out

Yes objective data collection is important to quantify what area has activity and when it occurred.With many data collection locations object tracking using time and location markers is possible.
An example application would be a number of units located in and outside the home (located at known rodent traversal points).It would be possible to track the rodent movement over time and even the direction. E.g. outside unit 1 (next to compost) detected rodent at time 1 then soon after, unit 2 (inside house next to pantry) detected rat activity. This movement pattern indicates the individual rat is likely to live in compost and then visit inside. Information like this leads to improved targeting of eradication methods.

Reply

Brent Beaven Oct 6, 2017

Hi Gerald.
Great to see that you have had good engagement and lots of questions on your submission.
I appreciate you taking the time to outline the pros and cons. I think that added a wealth of richness to your submission.
My advice now would be to take all of the feedback and questions and extra thoughts and refine our submission to incorporate this.
Get as clear as possible on what bits are a new innovation and how you are combining these with existing product (albeit in an innovative way) to achieve a working project. The more detailed and specific you can get will help the judges in understanding the value of this innovation.
Will check in again mid-week.
Good luck,
Brent

Reply

Gerald Dickinson Oct 7, 2017

Hi Brent,

Can I have some clarification on "...and how you are combining these with existing product"
When mentioning existing product which ones are you referring to?

1) Is it current invasive animal monitoring tools such as chew cards and tracking tunnels in use by the general conservation/predator free community trapping groups?
2) Is it human detection using thermal imaging
3) Is it other project, if so please specify which ones

Thank you for your clarification
Gerald

Reply

Brent Beaven Oct 12, 2017

Hi Gerald.
Apologies for the delay in responding. I was simply, at the time, encouraging you to be clear in what parts of the you have developed in an innovative way. I poorly used the term product, to describe any widely used technology that you have brought together in a new way, versus any components of your entry that you have invented. Apologies for poor use of language.
That said, I see that there has been a broader discussion on how this could integrate with other devices. Lots of potential.
Good luck with the entry,
Brent

Reply

View all replies (2)

Kevin Bain Oct 9, 2017

Hi Gerald, We (NZAutoTraps) have got the entry “One self-resetting trap that kills both Possums and Rats”.

Here’s a thought: Our trap is battery powered and triggered by photo eyes (electrical signal) so we can control when or if our trap was to trigger. If your black box could reliably differentiate between a pest (any pest) and a non-target, say a bird, then output an “OK to Kill” or “Don’t Kill” signal to our trap we could act accordingly. That is: if you see a pest approaching the trap you tell our trap “OK to Kill”, we enable our trap to kill when ready. If you see a non-pest approach you tell our trap “Don’t Kill”, we don’t enable our trap and don’t kill, we might even add a sound to scare the non-pest away??

Alternatively to increase your standby time (reduce power consumption) we could detect something using our low power IR photo eye and “Wake up” your controller, you Identify the target and echo back “OK to Kill” or “Don’t Kill”.

The big advantage this would have is now our kill trap designs could be more open and attractive to pests. No more little wee holes and complicated entries currently used to separate target from non-target.
Food for thought??
Thanks Kevin

Reply

Gerald Dickinson Oct 10, 2017

Kevin,
Yes adding the black box (grid-i) to identify animal in the area and prime trapping system for targets is possible. Combining the two pieces of technology is the way of the future and makes for a more flexible product range.
Will get in contact.

Appreciate your openess
Gerald

Reply

Joanne Jackson Oct 12, 2017

Cool idea. Will the cameras be placed just on the ground, or in other positions?

Reply

Gerald Dickinson Oct 12, 2017

Hi Joanne,
Thanks for the question. Initially the unit will be on a pole of fixed high above the ground, the benefits of this method include;

1) Quick and easy for on-site setup, Use the idea that farmers setting up temporary electric fences, require light posts with prongs therefore, easy to push in and pull out)
2) Easy to carry , light weight strong fibre glass poles, many can be carried into remote sites.
3) Thermal detector is a fixed height above. This provides a heat signature that is standardised size wise. Also simplifies the complexity of AI to match unknown heat signature to the internal library (template matching)
4) Protects thermal camera from elements. By mounting sensor downwards operation is not compromised by direct sunlight and rain/ water on sensor lens.
5) Animal abuse is minimised. Rats like to chew so having unit mounted high up above ground reduces this issue.
6) Antenna high above ground. Improves the signal reception range
7) Easy to setup in open areas where there are no natural mounting objects.
8) Don't need to fasten unit to tree (no damage to trees)
9) Easier to find when removing from site.

These are some advantages of having a standardised mounting platform. It is not recommended to mount each unit to tree or object as each location is different so creating reliability issues in detecting and identifying pest.
Thanks
Gerald

Reply

Chris Sissons Oct 12, 2017

Hi Gerald
Your combination of species identification , monitoring long-term , and distant long-range data-logging provides a potentially power resourse in many areas.
Beyond predator ID and use in monitoring I see an important potential use in research to where vertebrates may be important in transmission of seeds and diseases - in native forests, in fact in any land ecosystem of interest.

Springing to mind, for example, the situation in Kauri forests where Kauri dieback is an urgent and major possibly extinction problem for Kauri. The causitive Phytophora agathidicida (PTA) is soil resident and spread by soil transfer such as on human feet. DOC has spent millions on boardwalks and wash stations on major tracks with limited success , hence the Auckland Regional Council proposal to completely close thre Waitakere bush.. Part of the problem is that coverage is limited, not all tracks are boardwalked, etc and it would be helpful in management stategies to monitor unregulated off-boardwalk pedestrian traffic in key areas,. But in addition to humans, other animals potentially transmit the pathogen. An Auckland University study on transmission by feral pigs, by examining soil in trotters and snouts, was suggestive but inconclusive. Smaller vertebrates are also potential vectors. Their activity around Kauri not well known. Using your device for long-term remote monitoring and identification of mobile species in senstive areas would potentially help greatly to understand and control and management of Kauri dieback vectors.

This device seems to be a very clever product utilizing new powerful tecknologies and uniques species recognition alogrithims with a very wide range of important uses.

Reply

Gerald Dickinson Oct 12, 2017

Hi Chris,
Thank you for your feedback. There are many future applications and pushing the technology boundary to aid this work is what I enjoy most. Having community feedback and support from scientists and others is key to achieving this goal.
For example having support from Landcare Research in Lincoln for field trials to collect heat signatures on invasive pests (rats, stoats, possums) is key to determine what makes animals unique in the heat spectrum. In the future other species (feral pigs) can be added to the library and expand the application range.

This technology was trialled on humans as a smart pedestrian monitor at Otari-Wilton bush in Wellington and showed this technology and application can detect larger heat emitting objects. See picture of prototype 1 in-field trial.

Thanks
Gerald

Reply

Joanne Jackson Oct 12, 2017

Hi Gerald. How will the data be retrieved, disseminated, stored and interpreted?

Reply

Daniel Bar-Even Oct 13, 2017

Hi Joanne, Gerald. Really like this idea and I think it definitely has legs. You could incorporate the data into Trap.NZ right now - just create a new monitoring station type and you can start adding records, load images, etc). I think Predator Free Ngaio is already recording data in there so you can overlay the monitoring records with their trap records and start getting some really useful and interesting data. If you want to look at feeding data in directly (i.e. add records and images via the API) then please get in touch - we are currently working on integrating other monitoring data from remote sensors so now is a good time!

Reply

Gerald Dickinson Oct 13, 2017

Hi Joanne,
Grid-i is about data capturing and transferring data from the many sensor nodes (private network) directly to a Grid-i base station via LoRa. (The size of the data record is small e g 20 - 30 bytes so volume is small)

The base collects the data and transfers to a central repository managed by a third party. This link from base to central repository can be done by
1) using the LoRaWAN network (under Telco control) often require subscription charges.
2) inside private home (piggy back base to network via WiFi network) private home owner is already paying for internet. (This is ideal for urban trapping communities eg predator free ngaio)
3) Use multiple base stations to create a network to a private home owner.

As to how the data is collected and managed in a central repository (Daniel Bar-Eleven commented on Trap NZ as an ideal candidate). It is then up to how the third party disseminates the data and how it provides access to the data. This is a huge fruitful area and would require plenty of resources to build an information management tool. There is plenty of places already investing in this area.

Gerald




Reply

View all replies (2)

Michele Frank Oct 15, 2017

The idea has been progressed to the next milestone

Reply

Becky Wilson Dec 3, 2017

Status label added: WINNER

Reply

Share