discussion / Wildlife Crime  / 1 July 2016

Ivory ID using portable molecular sensors?

I was recently introduced to portable molecular sensors at a conference. One example is Consumer Physics' SCiO. Does anyone have experience using portable molecular sensors? If so, I'd value input on whether this tool might be applied for low-cost, in-situ identification of 'ivory' in legal/illegal retail.

When 'ivory' is visually examined, it can be challenging to confirm whether the material is elephant ivory, especially if Schreger lines are not visible. Common elephant ivory look-alikes include:

  • Mammoth ivory
  • Walrus tooth
  • Sperm whale and killer whale teeth
  • Narwhal tusk
  • Hippopotamus tooth
  • Wart hog tusk
  • Animal bone and antler
  • Vegetable ivory (tagua nut)
  • Giant clam shell
  • Synthetic materials and resins

I assume all of these materials would have unique molecular compositions. Could these be identified and classified using a portable sensor? If so, it could have invaluable applications. Inputs appreciated.

There are so few details on the link that it is not possible to judge whether that particualr hardware will be any use for anything. I suspect that it used near infra-red which has some applications for bulk anayses of foods but I doubt that it could distinguish different types of animal ivory. It might be able to discriminate between vegetable and animal ivory.

Thanks, @Peter+Apps . I reached out to ConsumerPhysics for more information. They said that scanning animal ivories for identification purposes is likely to be feasible with SCiO and that others may be considering this application as well. 

They confirmed that SCiO uses NIR spectroscopy for classification between different substances, providing that the molecular fingerprints are different enough from each other. It sounds like they did a basic feasibility test that "showed promise in classifying ivory samples from different kinds of animals - elephant, mammoth, hippopotamus and walrus, as well as plastic mock-ivory samples. The corresponding molecular sensing models and applications will have to be developed further for better insight. This may be developed in the future either by us or by the SCiO developer community using the SCiO Development Toolkit, which allows users to create models and apps." 

They suggested that a user would need to take two things into consideration when working toward this application: "First, as SCiO’s detection threshold varies by material, and is typically 0.1%-1%, some micro-nutrients will not be detectable. This requires further research. In addition, it should be noted that building the database for such an application requires access to samples and their chemical characteristics."  

More info on DevKit SDK is online here: https://www.consumerphysics.com/myscio/developers/ and here: https://www.youtube.com/watch?v=6wDLSF0ZkIA. I have no experience with this technology and no connection to ConsumerPhysics, but it sounds interesting.

I think this could be a good research project for someone in the WILDLABS community. With the scope of our network, gaining access to samples should not be a challenge.

It sounds like @David+Baisch at Conservation X Labs just ordered a SCiO spec and developer's kit to test its sensitivity. They're interested in seeing whether the device can be used to differentiate beteween tree species in treated and untreated wood products. Looking forward to learning more about that process!