This community is in archive. Visit community.xprize.org for the current XPRIZE Community.

What emerging technologies can be used to help us more quickly identify new species?

XPRIZEXPRIZE Posts: 193 admin
edited October 2019 in Key Issues
70% of plants with anti-cancer characteristics can only be found in the tropical rainforest, yet less than 10% of tropical rainforest plant species and .1% of animal species have been examined for their chemical and medicinal value.

At the same time, the UN’s 50-year study - completed this month by the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services - demonstrates that nature’s decline worldwide is unprecedented and that species extinction rates are accelerating. Its findings demonstrate that the world’s massive loss of biodiversity is “eroding the very foundations of our economies, livelihoods, food security, health and quality of life worldwide.”

What emerging technologies can be used to help us more quickly identify new species?

Share any links, examples, or ideas that you have!

Comments

  • marz62marz62 Posts: 122 ✭✭✭
    Normally, I have references to back up comments, but I am not in my library right now to look up the article I am referencing (source: American Scientist, possibly 2015 or 2016). In any event, the article featured a type of (aerial) spectrographic analysis of rain forest tree cover that produced different spectral signatures (color variations) for different species of tree. However, this technique worked for identifying tree species via the top of the forest canopy -- a canopy that is often too dense for such (aerial) spectrometers to penetrate to the ground level where rare plants may be found.

    Now, the technology known as LiDAR (light detection and ranging) that uses super fast laser pulses that can indeed penetrate to the ground level (it is used to identify archeological ruins beneath dense forest/jungle cover) but does so by digitally 'eliminating' the vegetation cover (simplistically, by using a tuned laser system that 'extracts' anything green from the image). Since most plants are green, this poses a problem: how to detect ground plants from the air and also by-pass the forest canopy that obscures them?

    Perhaps a 'one-two' punch' type tuned laser system (Lidar + Spectrometry/spectrography) might be usable, where the first (lidar) laser pulses by-pass the canopy and a second (broadly 'tuned') laser pulse (stream) acts as a spectrometer or spectral 'probe' (e.g., Raman spectroscopy, or, Coherent Anti-Stokes Raman spectrometry, known as CARS) of the ground level vegetation (to identify new plant species or potential sites of possible new species). Lidar systems are now being mounted on drones for close-in, more precise, mapping purposes.

    All that said, any type of spectrometric /spectrographic tech may likely only be usable on the ground ("close up" as it were) due to the much smaller size of plants verses tree top/canopies. Further, precision tuning would be paramount as no doubt many species of plants will have similar (or overlapping) spectral signatures. This is why, in a separate forum post, I recommended use of portable DNA sequencers (e.g., MinION, or, Nanopore technology), which could be augmented by wi-fi capability (provided by aerial wi-fi-transmitting balloons or drones connected to the an Internet/cloud database of plant DNA sequences). Cellular networks could also be employed (in conjunction with a camera-spectrography app) -- assuming that cellular microwave transmissions can reach into remote forested areas.

    Lastly, I also recommended (as several others have) forging contacts with indigenous peoples/tribes (known as "stakeholders") to both identify their traditional medicinal plants, and, locate areas rich in plant diversity. Cultivating this type of cooperative relationship (in which their native land is protected and their way of life is preserved as much as possible, from outside interference) is paramount. This is a "low tech" approach that should never be over-looked in our rush to find and apply new technological solutions.
  • akbakb Posts: 212 ✭✭✭
    This presents a fantastic opportunity for an XPRIZE! Think about the challenges we have faced and overcome by sending autonomous rovers to the Moon and other planets (Mars). How about a challenge for (semi-) autonomous "rovers" that explore remote regions of the Earth (rain-forest, jungle, oceans, etc.)?

    These would be much more capable in their observation and chemical and biological analysis. Various methods of movement could be employed depending on the environment (e.g. wheels, tracks, crawling, climbing, flying, and diving).
  • marz62marz62 Posts: 122 ✭✭✭
    edited May 2019
    @akb - Yes, for sure! Crawling robots (which have ML programs to avoid trampling plants, avoiding obstacles, and self-correcting behaviors [see: Lipson and Zalaga, 2012?]) would be super useful. Again, we should always conduct 'outreach' with known indigenous tribes peoples, so as to alert them about the robots in their territory.
  • akbakb Posts: 212 ✭✭✭
    Good point @marz62
  • Kathleen_HamrickKathleen_Hamrick Posts: 66 XPRIZE
    @marz62 and @akb thanks a bunch for this insight and great ideas! Another question for both of you: what challenges (and opportunities) exist for innovators to layer information coming from different remote sensing platforms and sensor types? How might they combine data from different sources to gain new insights that can be used for good, or to make predictions? @TerryMulligan may have some additional thoughts/questions.
  • TerryMulliganTerryMulligan Posts: 38 XPRIZE
    First, I want to second Kathleen's thanks to you both @akb and @marz62 for your comments! We also feel like this is a great opportunity for XPRIZE!

    Regarding indigenous communities - we absolutely agree. Indigenous communities are our present-day and future stewards of the forest.

    We are very intrigued by the insights that can be gained by high-resolution layered data (i.e. Lidar + Spectroscopy). I believe the example you are referencing is from Greg Asner and the Carnegie Airborne Observatory mapping the Peruvian Amazon. It was very inspirational work, and it has us questioning - what else is possible?

    In addition to @Kathleen_Hamrick's questions - have either of you (or anyone else in the community) seen any examples of integrated remote-sensing technologies being used for "close up" observation? UAV mounted, handheld, rover mounted? Which combinations could have the greatest value?

  • marz62marz62 Posts: 122 ✭✭✭
    @Kathleen_Hamrick @TerryMulligan @akb Ok, that's a cool 'challenge' (within the Challenge). I will need a bit to get back to you on this...I am actually working two other challenges right now, one for ESA. Your question definitely overlaps with the research I am currently doing for both challenges.
  • akbakb Posts: 212 ✭✭✭
    Hi @Kathleen_Hamrick There have been many projects across different sectors that have addressed the challenges of integrating different data sources. No doubt some new challenges could emerge for this topic. A review of projects and mainstream systems using GIS, IoT and/or communications (e.g. traffic and transport information) could be useful. Also a review of information types used by professionals in this sector would be useful.
  • marz62marz62 Posts: 122 ✭✭✭
    @akb @Kathleen_Hamrick - Are we talking about integrated data sources...or, remote sensing technologies? Data is always good, but I presumed that this forum topic is being broached because of a lack of data (?)...thus the need for remote sensing technologies (to acquire said data)

    The use of GIS is certainly valuable, but its use assumes prior investigation /exploration/mapping of the terrain (e.g., dense forest) to some degree. If the terrain variation (i.e., different habitats) is encoded/defined in a given area, this could provide cues as to where to look for new species. Also, if prior expeditions to an area resulted in one or more discoveries of medicinal or rare plants, then this information could be quite useful. But absent any prior mapping of these kinds, GIS technology would seem only to be useful AFTER a terrain has been explored or had its flora 'inventoried' (whether via remote sensing or field study).

    As for IoT ...this is the hot topic these days but I am not sure how useful in would be in this context (apart from centralizing in situ sensor networks...which assumes we have targeted {identified] something already -- presumably via remote sensing tech).

    Please give practical or theoretical examples of how IoT would be used in this context.
  • akbakb Posts: 212 ✭✭✭
    Hi @marz62 and @Kathleen_Hamrick
    Good points.
    Projects might use in situ sensor networks for their IoT, and/or mobile IoT (like the rover XPRIZE idea I suggested). More broadly there are lots of technologies that already exist to collect data and more are emerging. A key aspect for success will be data compatibility and the need for a common standard*.

    * Have you heard the old joke? "Standards are a wonderful thing, and we should have lots of them" [too many] ;-)
  • NickAzerNickAzer Posts: 219 ✭✭
    Thank you @akb and @marz62! The GIS question is interesting. @TerryMulligan and @Kathleen_Hamrick, do you have any further input?
Sign In or Register to comment.