On September 14, 2019, the state-owned Saudi Aramco, one of the largest oil producers in the world, was attacked by several unmanned aerial vehicles (UAVs), causing panic, damage, and a spike in global oil prices. Although the facility attacked had several defense systems, all of them proved ineffective for a low-flying drone.

On a smaller scale, faculty at the Eversource Energy Center at the University of Connecticut are working hand-in-hand with its namesake to build a drone detection system. One of the people leading the project, Chandi Witharana, assistant research professor of natural resources and the environment, explains the plan.

Q: Why is a sensor-based surveillance system a good idea for a utility company? 

A: We should take Aramco as a loud ‘wake-up’ call, which we never heard before and we pretty much overlooked. Even a short-run power failure mediated by a drone attack to electrical infrastructure comes at a high price.  On top of that, such failures can trigger cascading events in the grid leading to much larger scale blackout. Imagine a coordinated attack on multiple substations, which would cripple the social and economic operations. Critical utility infrastructure have surveillance systems for perimeter instructions etc. Threat from air is a new form, which requires the same level of attention and surveillance as ground-based threats. As long as people have access to drones the threat prevails.

Q: How vulnerable are electrical sub-stations to drone attacks?

A: The risk prevails in all directions despite the security measures. Although the drones have been made available to civilians, most of the detection, tracking, and aversion technologies are still being confined to military use. Access to the few number of available detection technologies are mainly prohibited by the cost. To give a sense, some advanced systems costs in the range of $100k to half a million dollars. Drone detection is also a growing market but still way behind compared to the drone market. In a sense, the drone technology is disruptive. It has not settled down yet.  We have so far been seeing the bright side, however, the climate is quickly tuning for more of ‘misuse’ of drones worldwide given growth of the market.

Q: To install a detection system, how much time would it take to implement, and how much work would go into an installation?

A: Commercially-available detection and tracking devices are well-packaged and readily deployable, however, they are prone to limitations in terms of detection and tracking. What we are exploring is the detection capabilities cost effectiveness of commercial sensors with respect to drone types, movements, and varying environmental conditions. One of the key aspects we are looking into is how to analytically combine the data streams coming from multiple sensors to make a real-time prediction of intrusive drones and their paths. Collectively, we want to develop and test an early warning UAV detection system using commercially-available devices to detect and track unauthorized UAV operations near electric substations.

Q: Tell us a little about the work Eversource Energy Center is doing around using sensors for surveillance of electrical substations. What spurred this work?

A: While drones are radically becoming  ‘everyday’ technology by reshaping the boundaries of scientific, industry, and business domains, their concealed threat to human security is on a steep rise due to the high probability of misuse of the technology. It has been predicted that during the years 2016 to 2020, the drone technology market will exceed $100 billion.  This simply means that we have access to drones more than ever before from recreational to commercial use. Drones are getting smaller, cheaper, ubiquitous, and, today, anyone can deploy an off-the-shelf amateur to professional-grade drone using a smartphone app without being a tech savvy. These virtues make appealing tool for use by individuals for nefarious purposes.

Surveillance systems are advancing in multitude directions. Miniature drone and light-weight imaging sensors — not only standard RGB camera but also thermal — can easily be used in reconnaissance of potentials targets. For example critical infrastructure such as electrical substations. Captured overhead imagery could help locating sensitive equipment inside the substation, and aid in planning a subsequent attack mediated by another drone via a direct crash or outfitted with explosives. Navigation systems of systems of drones have substantially been improved by allowing accurate autonomous flights pre-planned areas. Another direction of improvement is the payload capacity — the maximum weight can be vested to the drone. Nowadays, professional-grade drones have the capacity of flying with couple of kilograms of additional weight. These drone aren’t so expensive either. Typically in the range of $5,000. One could easily get the advantage of payload to attach explosives to make a targeted attack on an electrical substation.  Parallel to payloads, the market has targeted and automated payload release devices that are designed to use in fishing and other applications. Using such devices, the explosives can be released with a high precision on to the electrical equipment.

Although we have not witnessed a drone attack yet in the U.S., the technology has grown to be a threat to our electrical infrastructure. In the report of “U.S. Critical Infrastructure 2025: A Strategic Risk Assessment,” the Department of Homeland Security emphasized the profound effect of aerial threats from drones on our critical infrastructure.  This topic is especially important to Eversource Energy and our center because of the potential risk on electrical substations in the Connecticut.

Q: After the implementation of a surveillance system, how much work and personnel would be needed to monitor for possible threats? 

A: Once the integrated sensor system is developed, we expect to establish an automated  alerting mechanism that sends out message to Eversource officials, which pretty much provides the information on the intrusive drone, potentially the location of the operator, flight path, and other salient details dynamically updated on a map.

Q: Why is it important for UConn researchers to work on this, with the facilities at UConn Tech Park, rather than Eversource working on this in-house?

A: Our key strengths in remote sensing science and technologies and advanced analytical capabilities allow us to integrate multiple sensing mechanisms to accurately detect and track drones. Our research will identify the advantages and limitations of existing technologies under various conditions in a substation environment and develop integration software to help improve detection accuracy and facilitate alerting Eversource personnel.