After having deployed to as many as six locations in Africa and the Middle East, personnel associated with the Defense Department\u2019s hallmark automation effort say they are learning valuable lessons in how to use algorithms for war.<\/p>\n\n\n\n
Known as Project Maven, the DoD initiative aims to accelerate the integration of big data and machine learning, first focusing on processing of full motion video from tactical drones and from medium altitude sensors.<\/p>\n\n\n\n
Here\u2019s what officials are saying:<\/p>\n\n\n\n
It\u2019s critical to retrain algorithms aimed at analyzing full motion video<\/b><\/p>\n\n\n\n
\u201cIt\u2019s not all that different if you think about it than a young airman coming onto my ops floor. On their first day of work it\u2019s going to take them a little while to figure out what\u2019s going on, to understand the mission, to understand the [area of responsibility] in which we\u2019re operating,\u201d Lt. Col. Garry \u201cPink\u201d Floyd, deputy chief of the algorithmic warfare cross functional team, also known as Project Maven, at the Modular Open Systems Summit in Washington May 1. \u201cYou see a similar version of that with [artificial intelligence], with these algorithms.\u201d<\/p>\n\n\n\n
Analysts are currently swimming in troves of data that are nearly impossible for humans to sift through.<\/p>\n\n\n\n
To exemplify the problem, military personnel at the annual GEOINT symposium in Tampa, Florida in late April said in fiscal 2017, 127 terabytes of captured enemy data was collected. In addition, the annual video that Central Command collects could cover 325,000 feature films and the annual signals intelligence Central Command collects is equal to roughly 5.5 million songs.<\/p>\n\n\n\n
Floyd said in Africa Command where the first algorithm was deployed, staff retrained the machine six times in five days.<\/p>\n\n\n\n
\u201cMaybe they\u2019re trained off of data from one region and we deploy it to a new region and maybe makes a few silly mistakes at first, but we\u2019ve developed some tools to help do that quickly,\u201d he said, noting these algorithms have been deployed to garrison processing, exploitation and dissemination sites.<\/p>\n\n\n\n
Floyd explained the team built into user interfaces a button that literally says \u201ctrain AI.\u201d<\/p>\n\n\n\n
\u201cIf you see the algorithm misidentify a palm tree as a person or something like that, the analyst, the operator can hit that train AI button, it captures the number of frames to the left of that instant, it captures the frames to right. Take that out of the theater and get it into our data labeling pipeline,\u201d he said. \u201cWe relabel the data, we rapidly get that to our algorithm developer \u2026 they retrain optimize \u2026 and then we redeploy it to the field. We were thinking this would take a long time, but again, we were able to do this in like say a day or so or even less.<\/p>\n\n\n\n
Data labeling is a necessity<\/b><\/p>\n\n\n\n
The algorithms must have labeled, or what is often referred to as structured, data to work with as opposed to raw data or unstructured data. Without it, the algorithms struggle to complete their work.<\/p>\n\n\n\n
Floyd said one approach to the data labeling issue is grabbing airmen that are waiting on their security clearances to do the labeling.<\/p>\n\n\n\n
\u201cSome of our best pockets of data labelers are folks that are waiting orders or waiting to get their clearance and it\u2019s just been tremendous,\u201d he said. \u201cThat becomes their service to their country for those first three or four months while they\u2019re waiting to get in the front door at Fort Meade or some other place like that.\u201d<\/p>\n\n\n\n
Adaptive interfaces<\/b><\/p>\n\n\n\n
User interfaces for the operators and analysts leveraging the algorithms must be tailorable and adaptable, Floyd said.<\/p>\n\n\n\n
\u201cWe need to give the user interface to the operators to where they can choose the algorithm they want to use for the mission that they\u2019re on. An algorithm that\u2019s tailored for this area of the world won\u2019t be perfect for another area of the world,\u201d he explained. \u201cWe know that we\u2019ll have tailored algorithms for different geographic regions.\u201d<\/p>\n\n\n\n
One of the most important aspects of deploying algorithms in conflict is going to be how operators use it, Floyd said describing a \u201cslide rule\u201d of confidence regarding the algorithm\u2019s performance.<\/p>\n\n\n\n
Some operators might want the algorithm to cue things that it\u2019s 80 percent or higher in, he said. Others might say slide the confidence down to 20 percent because that might return hits the human wouldn\u2019t normally see.<\/p>\n\n\n\n
\u201cWe have seen that a couple of times already in the deployed environment where the algorithm picked up things the human eye missed,\u201d Floyd said.<\/p>\n\n\n\n
Future of Project Maven<\/b><\/p>\n\n\n\n
Floyd said DoD is still in the early stages of Project Maven adding that while the technology has been deployed in six locations, more are on the way. There\u2019s a pathway for additional sites in the U.S. and overseas throughout the rest of the year.<\/p>\n\n\n\n
He alluded to additional use cases or applications beyond full motion video for Project Maven.<\/p>\n\n\n\n
\u201cIt might be a little early to get too specific on those. But there might be another use case or two, another sensor type or two that we might go after soon,\u201d he said.<\/p>\n","protected":false},"excerpt":{"rendered":"
Project Maven has already deployed in five or six locations across Africa and the Middle East. Here’s what officials are learning from those deployments so far.<\/p>\n","protected":false},"author":7,"featured_media":37632,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"_yoast_wpseo_meta-robots-noindex":"","_yoast_wpseo_meta-robots-nofollow":"","_yoast_wpseo_canonical":"","_acf":"","_yoast_wpseo_primary_category":41,"_jetpack_feature_clip_id":0,"_jetpack_memberships_contains_paid_content":false,"footnotes":"","_smg_distribution_targets":[]},"categories":[106,29,15,41,32],"tags":[],"coauthors":[3268],"class_list":["post-23102","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-daily-brief","category-home","category-intel-geoint","category-isr","category-newsletters"],"acf":{"subheadline":"","legacy_arc_id":"ZK7UP6ZIWFACRFHETB7O2Z7O4Q","arc_canonical_url":"\/intel-geoint\/isr\/2018\/05\/02\/what-the-pentagon-is-learning-from-its-massive-machine-learning-project\/","remove_feature_photo":false,"is_sponsored":false,"subtype":"","redirect_url":"","disable_inline_ads":false,"native_logo_pretext":"Presented By:"},"yoast_head":"\n