What’s the One Thing Stopping Drones From Taking Off in Agriculture?02/02/2017
Unmanned aerial vehicles (aka drones) have received a lot of attention in the precision ag sector recently. AgFunder reported in 2015 a total of $389 million was invested in drones and robotics companies representing 8% of the $4.6B total investment in AgTech that year. Despite the attention drones have received from investors and the media, it seems some of the shine has begun to wear off.
From discussions with growers and precision farming dealers across the US, it appears drones have flown into the trough of disillusionment. A lot of the folks I’ve talked to say their drones are being shelved for now.
So what is the headwind preventing take-off for drones?
In one word…
Despite drone technology advancing rapidly and commercial hardware becoming more affordable, there still remain significant obstacles to overcome. Obstacles such as poor operational efficiency, limitations in software processing and complexity in turning raw data into real-world decision support outcomes for growers. The commercial payback is yet to be realised for mainstream drone operation.
Don’t get me wrong, there are early adopter growers and business operators that are making a go of it. But there are still a lot of steps required to make it work.
High friction workflows do not scale for the masses.
Nevertheless, I think there is a bright future for drones in agriculture once these issues are addressed and a suitable level of autonomous deployment and data automation is achieved.
Obstacle #1 — Friction of hardware deployment
Currently there is considerable time (and therefore cost) required to map a field. The time required to drive to the field, establish a flight plan, deploy the drone, monitor its performance and download the imagery.
Today a commercial drone has a flight time of around 40 minutes which equates to approximately 100–200 acres depending on a number of factors including:
- image resolution
- image overlap required
- wind speed
- platform configuration ie quad copter versus fixed wing
Configuration of drone platform will also determine how you might use the drone due to vertical take off and better manoeuvrability of quad copters. However, it’s difficult for drones at any level to compete with the coverage efficiency of aerial and satellite systems.
Solution #1 — Autonomous operation
Once drones are able to take off autonomously, scan a field, upload data, recharge and continue operation without human intervention — this will be a game changer. Continuous automated image capture eliminates friction from the raw data acquisition. It’s like having your own private satellite system orbiting your farm. Every day!
Images from across the farm would be streamed over Wi-Fi to the farm PC and then uploaded to the cloud for processing and analysis. Integrated weather data would enable the drone to know when to return to base or prevent take-off during poor weather conditions. Areas of interest would receive extra attention by increasing frequency of coverage and flying at lower altitude to increase image resolution during critical growing periods.
Boom! Frictionless and continuous image capture. The same as what is already occurring with agronomic data automatically streaming from ground based tractors, sprayers and harvesters.
Note — There are also other barriers to overcome to achieve this level of autonomous operation such as changes in legislation to allow beyond visual line of sight (BVLOS) operation. But that’s a whole other debate.
Obstacle #2 — Friction of software processing
With the autonomous take off, landing, battery recharging and image streaming to the cloud then the foundation is laid for automated data processing. Raw images automatically uploaded to the cloud by the drone can then be stitched together, analyzed and interpreted for enhanced decision-making.
Consider the friction that was removed from personal photography during the transition from film to digital processing. In the “old days” you took a photo on your polaroid camera, waited until the film was full and dropped it off to get processed at the lab. Days (or weeks) later the photos were ready to collect. Occasionally excitement turned to disappointment only to see your favourite happy snaps were out of focus or your finger was over the lens!
These high friction issues are now a thing of the past with the introduction of digital photography. The real-time feedback (point, shoot, view) and zero cost of production means that you continue to take photos until you’re are happy with the result.
Autonomous drones must follow a similar path.
Solution #2 — Automated Image Processing
Drones manufacturers and software providers must remove the friction of “sending images to the lab”. With autonomous hardware that’s able to take-off, record and upload images in real time — processing of those same images must also be intelligent and automated. Wireless delivery and integration of the processed results into farm management software platforms is also critical to maximising drone value. Companies such as AgDNA are already performing similar automated data processing and analytics with vast amounts of agronomic data from ground based vehicles.
Recapture of images might also occur if some are of poor quality or if there is a region of interest that requires further investigation. The drone would automatically return to a particular hot spot and fly at a lower altitude to take higher resolution images for disease or pest identification. The grower or agronomist would then be advised as issues are identified without even stepping foot in the field.
As with all tools if you’re unable to use the information to make a decision for a particular outcome then the value of the tool is questionable. This has been the case with drones for some time and is the reason why they have entered the trough of disillusionment. However, once a level of autonomous operation outlined above is achieved the outlook for drones will improve significantly.
Ultimately drones will be continuously monitoring the field without the need for humans – all at very low operating costs. Measuring plant stand count after seeding, identifying drainage issues after heavy rainfall, checking for pests and diseases, measuring biomass and forecasting potential yield can all be automated using drones.
All of this valuable information can then be used at the point of highest impact during the season when decisions and action can be taken.
The drone industry has been in a holding pattern for sometime. Waiting for FAA regulations to be outlined. Commercial operation of drones is now legal however there are still restrictions such visual line of sight (VLOS) operation that prevent automatic deployment of drones. It’s hoped that regulations will continue to evolve (eg BVLOS operation) and extract the real value drones can deliver for agriculture.
Once the legal pathway is clear the rest is an engineering problem-solving exercise. It’s fair to say that the technology required to automatically fly, capture and process imagery is already available. Now it just requires an innovative AgTech company to connect the dots and make it possible.
Once the FRICTION is removed then drones will come out of the trough of disillusionment and enter the slope of enlightenment.
Drones are destined to deliver on their promise that so many growers and precision farming experts are seeking.
Let’s hope the technology and required legislation arrives soon so growers can begin unlocking the real potential economic value they offer.
Source: Paul Turner, CEO at AgDNA on LinkedIn