US Army tests drones that sniff out clouds of chemical agents
Drones are proving their worth in dangerous situations we don't want to expose humans to by keeping an eye out for sharks, mines and methane leaks. Now a team from the US Army's Edgewood Chemical and Biological Center (ECBC) have tested a new pair of unmanned vehicles to detect chemical and biological agents.
The S/K Challenge is an annual event held by the US Army that allows different agencies and organizations to test their chemical and biological detection technologies by releasing clouds of "simulant agents" in controlled environments at the Dugway Proving Ground. In this year's event, held from August 15 to 26, ECBC entered Deep Purple, a quadcopter drone made from commercial and custom parts, and an unmanned ground vehicle known as the Mobile Detection Assessment and Response System (MDARS).
To take advantage of the calmer wind and weather at night, the tests were held over those two weeks between the hours of 11pm and 5am. A simulated agent was released into the air somewhere within the test space and with the help of stationary sensors, the drones would be sent off to intercept the cloud and identify the agent.
The sensors used were the Joint Chemical Agent Detector and the Tactical Biological Generation II Detector (TACBIO), which specializes in detecting airborne biological threats. These were packed into thermos-shaped containers and attached to the roof of MDARS and the underside of Deep Purple, respectively.
Rather than carbon fiber, the drone's airframe consists of printed circuit boards, which the team says allows for direct, real-time communication with operators or field soldiers up to 2 miles (3.2 km) away. This data, along with that gathered by MDARS, is then collected by a central information sharing system.
"This is a system of systems, every part has to be able to communicate with every other so it works as an integrated system," explains Steven Lagan, a ECBC team member. "We get data on the location and movement of the simulant cloud from stationary sensors, which we can then send to our sensor mounted vehicles. The sensors then communicate to us the identity of the agent, which we share with all the other participants, and if this were real, the chain of command, through a common operating language."
Getting the various components to understand each other was a challenge, the team says, but some creative coding helped them through it. And in the end, the tests highlighted some other weaknesses of ECBC's systems, which can be focused on and honed before next year's event.
"We saw that the chemical sensors mounted on Deep Purple worked well at intercepting the cloud and identifying the simulants," says ECBC team leader, Alan Samuels. "However, our biological agent sensor, TACBIO, is too large for Deep Purple and only worked on MDARS. It needs to be miniaturized for that purpose. The S/K Challenge proved its value to ECBC by revealing what our system does well and where it needs more work. We'll come back next year having built upon what we learned this time."