InterDrone: a Delegate’s Perspective by Sol N’Jie

Sol N’Jie of Sol Management Services was the first to apply for a free ticket to InterDrone when we advertised this in our July Members’ newsletter. He kindly agreed to come back and write about it for us. Over to you, Sol

 

“The rapid shift from recreational drones taking photos and videos to commercial data gathering platforms feeding drone technology ecosystems was a consistent theme at the International Drone Conference and Exposition (InterDrone) last week.

InterDrone comprised of a conference program, exhibition (160+ drone manufacturers, sellers and solution providers), drone courses, panel sessions and networking events at what was billed as the largest, commercially-dedicated, UAV show. There were also many opportunities to meet a variety of drone pilots, service providers, UAS engineers and developers, UAV manufacturers, videographers and enterprise UAV end-users.

The key notes each day featured industry speakers from the FAA, DJI, 3DR, Parrot, NASA, Yuneec, Airmap, Scholar Farms and UNICEF. Also on offer were 6 specialised enterprise drone tracks in:

  • Construction, Mining and Aggregates
  • Precision Agriculture
  • Surveying and Mapping
  • Energy Inspection
  • Infrastructure Inspection
  • Public Safety

I was lucky enough to attend and represent ARPAS at InterDrone last week in Las Vegas. Although the InterDrone ticket was provided free I paid for all my own costs.

The striking parallels to the evolution of the Smartphone, personal computer and internet were all too evident over the three days, with drones having grown exponentially from a DIY hobby to a mainstream technology with flight features and camera quality that continues to improve at a phenomenal speed.

Autonomy, stability and reliable flight control systems are standard in commercial drones today,  greatly reducing the skills necessary to manage a drone operation. This has led to commercial drones becoming a valuable business tool, equipped with onboard sensors capable of high-level data collection, spawning many products and applications within the vertical tracks listed above.

Many of these applications rely on large scale data collection across many different and varying types of sites. Increasingly local machines cannot manage the massive amounts of information so storing and processing is moving into the cloud with the data then distributed throughout a business. The benefits of drone technology are starting to go beyond incremental improvements with a number of businesses transitioning from practices, typically conducted by personnel, to drone-based processes.

Artificial intelligence will open up many more opportunities in the near future. Drones will incorporate machine learning, automated feature recognition, change and defect detection, reducing the amount of data transfer, due to greater on board AI processing and work in swarms to achieve greater results. Once this is established, businesses can use this as a base platform for industry, application and predictive analytics to provide a business with a competitive advantage.

A number of challenges were also discussed that included creating the air traffic management systems that allow drones to operate at scale without interfering with each other or traditional aircraft. These issues will need to be resolved in order to unlock the full potential of the drone market. These challenges include:

  • Remote ID tracking with a licensed and unlicensed solution (8 technologies viewed as viable) however, no one solution can currently independently verify ID and location.
  • Collision avoidance (only one drone to date has all 6 directions of obstacle detection)
  • Operating beyond VLOS enabling deployment of drones over long distances
  • Creating the regulatory environment to allow autonomous flights
  • The management and feasibility of multiple or swarms of drones

So where are we in this great evolution? The following represent a few of the innovative solutions discussed during InterDrone.

Chris Anderson, founder of 3DR that provides drone software for construction, engineering, and mining teams, spoke of how construction is one of the least digitized industries in the world. “It’s ripe for transformation, but because it’s a large and a relatively conservative industry, it has been resistant. Today, that’s changing.”

In early 2016 3DR launched Site Scan, a drone data platform with flight and a web app for processing drone photos into maps and models and then analysing results. It offers a suite of tools for getting insights out of drone data, such as measuring distances and volumes, overlaying blueprints to compare design to reality and export data into other tools like Autodesk BIM 360, Revit, Civil 3D and InfraWorks 360. This helps teams, clients, and other stakeholders get real-time updates about construction projects to see whether a project is going as planned.

A more traditional use of drones was presented by the Roswell Flight Test Crew who drew on experiences using small, low-cost UAS equipped with thermal imaging cameras to assist firefighters and other first responders in a range of scenarios that include structural and wildland firefighting, search and rescue, as well as hazardous material spills. They covered the capabilities and limitations of thermal imaging technology, as well as its specific application in emergency response situations. UAVs are fast becoming a standard piece of kit for the emergency services and a number of innovative uses during emergency situations are available.

Michael Ott spoke about delivering targeted liquid sprays precisely where they are needed in agricultural applications due to decades of use (and overuse), which has led to chemical resistance and superweeds. Farmers are awash in data and need accurate intelligence to help them apply chemicals exactly where they are needed rather than blanket application to all areas. Ott’s company Rantizo, pairs GPS based drone technology to gather imagery on a field, then uses that data to diagnose problem areas and autonomously deliver chemicals, via an electrostatic spray on the drone, to target and reach specific areas within a field.

He described the current situation of having a gallon of Round Up that needs to be diluted with over 800 gallons of water and administered by a tractor because there is 801 gallons of liquid that needs spraying. Ott explained. “We don’t need that 800 gallons of water, we just need that one gallon. We can put our sprayer on a drone and spot apply where it needs to go. Precision in spraying will lead to more efficient use of chemicals and labour as well as reduced exposure to increasingly toxic agrichemicals.

Martin Fuchsberger and Benoit Fredericque covered the benefits for ‘Automated Workflows for Cell Tower Inspection’. Their opening statement ‘the world’s 3+ million cell towers are inspected on an annual basis anywhere between one and ten times traditionally by experienced climbers or more recently by manual or semi-automatic drones’ was a surprise.

Automated workflow for inspecting cell towers, is now available, that achieves the goals of data acquisition, 3D model reconstruction and model consumption and can be automated and streamlined to produce good, repeatable results at a dramatically lower cost than both climber-based and manual drone workflows. Additional advantages of using drones in this market are the dramatic increase in safety, reduced costs and increased inspection quality.

Counting Sheep with Drones was a fascinating example of how the integration of a number of technologies resulted in a drone AI based app that literally counts how many sheep there are in the farmer’s field. The team had previously developed an AI app for a drone to fly autonomously around a warehouse doing inventory checks by reading the bar codes and image recognition on the side of the boxes. To count sheep in a field they used  Google’s Tensorflow which can be used to relatively quickly add trained models to any Android app. They used machine learning in the Android app through a process of 5,000 individual pictures of sheep and herds of sheep to process the video and then trained the model, implemented the app and tested it in the real world. The ramifications and opportunities for these types of AI applications are huge.

InterDrone demonstrated that we are a part of a fast evolving industry that is becoming relevant to huge swaths of commerce and public entities. The future multibillion dollar revue tag, projected for the drone industry, seemed more realistic after spending three days at InterDrone.”

Solomon N’Jie    Sol Management Services