Can you build the best algorithm that predicts how the chemical spectrum of a substance changes in different molecular environments?

THIS CHALLENGE IS NOW OVER. Thank you for all your submissions!
For information regarding results of the Challenge, please contact:

July 2017 August 2017 September 2017 Fall 2017
Registration Opens
& Contest Begins
Build Your


Imagine if there was a way to identify an unknown substance in real-time, by simply pointing a scanner at the surface in which the substance is located. Chemicals, or areas containing potential chemical residues, could be scanned at standoff range and the substance identified on the spot. Such a development would enable expedited results in time sensitive situations and “one hundred percent” screening for security or process control. The Modeling of Reflectance Given Only Transmission of High-concentration Spectra for Chemical Recognition Over Widely-varying Environments (MORGOTH’S CROWN) Challenge seeks development of algorithms that would help accomplish just that!

This summer, Intelligence Advanced Research Projects Activity (IARPA) is excited to conduct the MORGOTH’S CROWN Challenge. The Challenge invites participants from around the world to develop algorithms to predict changes in a chemical’s infrared (IR) spectrum caused by changes in its molecular environment. Participants will join a global community working to benchmark research and foster innovation in IR spectral prediction through crowdsourcing.

The MORGOTH’S CROWN Challenge is related to our Standoff Illuminator for Measuring Absorbance and Reflectance Infrared Light Signatures (SILMARILS) program, which works to improve real-time methods for standoff detection and identification of chemical residues on surfaces using active infrared spectroscopy with spectral, physical, and chemical information to yield hardware and software models that are fast and accurate.


IARPA is conducting the MORGOTH’S CROWN Challenge to invite the broader research community of industry, academia, and individual code wizards to participate in a convenient, efficient, and non-contractual way.

IARPA’s use of a crowdsourcing approach to stimulate breakthroughs in science and technology also supports government transparency and efficiency. The goals and objectives of this Challenge are to:

  • Promote research and enable more comprehensive and robust detection libraries for both passive and active standoff infrared chemical detection
  • Encourage various communities to develop and enhance computational models to accurately predict spectral signatures of a given substance and substrate combination in complex environments
  • Foster innovation through crowdsourcing and move beyond current research limitations for algorithms and models enhancing infrared chemical detection
  • Cultivate and sustain an ongoing collaborative community dedicated to this technology and research

Participants will be given spectra of both clean surfaces (substrates) and chemicals of interest in their bulk form. They will also be given the spectra of a set of training coupons with different combinations of substrates with trace chemical residues on them (including different particle sizes, crystal structures, and mass loadings). These training coupons will show how the spectrum of a trace residue differs from both the bulk chemical and the substrate spectra due to the complex interplay at the level between scattered light and molecular structure. Participants are asked to generate an algorithm to predict what the spectra of DIFFERENT combinations of chemicals and substrates would look like. The evaluation phase will judge participants’ algorithms based on a number of accuracy variables related to prediction of the correct sample coupon spectra.

During the Challenge, solvers will work on their algorithm and track their performance against other competitors on a leaderboard. Participants will improve their algorithm throughout the Challenge duration.

"I don’t know, and I’d rather not guess."

– Frodo Baggins, The Fellowship of the Ring, J.R.R. Tolkien

  Prize Purse of $50k



1st Place ap31 Dmitirii Khrustalev
2nd Place ZFTurbo Roman Solovyev
3rd Place albantor30
4th Place cannab Victor Durnov
5th Place Tomasz Dyczek Tomasz Dyczek


MORGOTH'S CROWN Challenge will have two tracks: a prize track and a non-prize track. The non-prize track is open to SILMARILS Performers and Affiliates and other non-qualifying participants. Winners will be announced for each track, but only those winners in the prize track will be eligible for cash prizes from a total prize purse of $50,000.

First Prize Icon
First Prize


Second Prize Icon
Second Prize


Third Prize Icon
Third Prize


Fourth Prize Icon
Fourth Prize


Fifth Prize Icon
Fifth Prize


Bonus Prize Icon
Bonus Prizes


Winners from both tracks will be publicly recognized by IARPA in the final announcements based on their performance. Throughout the challenge, Topcoder's online leaderboard will display your rankings and accomplishments, giving you various opportunities to have your work viewed and appreciated by stakeholders from industry, government, and academic communities, regardless of the track in which you qualify.

For full participant eligibility requirements and official rules, visit


Participants are asked to develop a chemical spectra prediction solution, given target chemical identity, substrate identity and quantitative metadata. Sample spectra, with full sample metadata will be provided for training your algorithm. Each submitted solution file will be quantitatively compared to the ground truth (spectra measured using a real sample.) For each spectra provided, a score that indicates how well the predicted spectra matches the ground truth spectra will be calculated. Scores of multiple spectra will be combined to provide an overall score for each participant.

  • To calculate a quantitative matching score, curve comparison metrics will be implemented. These metrics will objectively quantify the difference between the two curves (the predicted spectra and the ground truth spectra).
  • In the calculation, metrics are normalized, allowing for the determination of a combined weighted average for each predicted/ground truth pair. An overall score of 1 indicates a perfect match, while a score closer to zero indicates a worse match.
  • The combined weighted averages of individual spectra will be shared with participants. Combined weighted averages of all spectra in the dataset will be combined to form an overall ranking score.


To assist builders in this challenge, IARPA has gathered a list of resources to prepare for each stage in the Challenge. You can find overview documents summarizing key parameters and physical models relevant to chemical spectral prediction, a wide variety of relevant literature references, information on how the challenge datasets were gathered, and detailed metadata for each of the challenge spectra to support your submissions.


The Intelligence Advanced Research Projects Activity (IARPA) invests in high-risk, high-payoff research programs to tackle some of the most difficult challenges of the agencies and disciplines in the Intelligence Community (IC).

IARPA LOGO Morgoth's Challenge
© IARPA. All Rights Reserved