fbpx

FarmSense: ‘Pesticide-free future with insect monitoring’

FarmSense makes, installs and maintains a digital monitoring system that alerts growers and eliminates uncertainty about the pests in their fields. The FlightSensor is trained to recognise and capture the wingbeats of each pest. - Photo: FarmSense
FarmSense makes, installs and maintains a digital monitoring system that alerts growers and eliminates uncertainty about the pests in their fields. The FlightSensor is trained to recognise and capture the wingbeats of each pest. - Photo: FarmSense

Smart pest management and insect monitoring could greatly reduce the use of expensive and damaging pesticides and even help lead to a pesticide-free future, says Dr. Eamonn Keogh, co-founder of FarmSense and professor at University of California, Riverside. However, there are challenges, many of them more cultural than technical.

FarmSense is a U.S. company that develops insect monitoring solutions for agriculture. The company is pioneering computational entomology to drive innovation in agriculture. By combining novel classification algorithms, Farmsense developed its proprietary FlightSensor, and can provide real-time insect counts for farmers, enabling faster access to critical data that can inform pesticide use and impact crop yield.

Data mining

Recently, Dr. Eamonn Keogh, co-founder of FarmSense and professor at University of California, Riverside, was honored with the 2021 IEEE International Conference on Data Mining Research Contributions Award – the organisation’s highest recognition for research achievements in data mining. Data mining is something that is mainly mentioned in the context of social media or advertising. “I am proud to have been honored for expanding data mining beyond social media and advertisements, to more socially important problems like understanding, and ultimately controlling, insects,” Dr. Keogh said.

How does FlightSensor work? What does a farmer need to know or do to be able to use it effectively?

“The farmer needs no special skills to install the sensor. They will typically place it on the same pole they have placed their mechanical traps for years, and use the same attractant etc. The FlightSensor shines an invisible (to both humans and insects) light been across the entrance of the trap. A sensor records the amount of light that crosses the entrance. It the light curtain is interrupted, by an insect flying inside, it changes the amount of light that is sensed in a principled way that is unique to an insect, a sort of ‘fingerprint’.”

What separates FarmSense’s FlightSensor from other systems?

“Almost all other systems simply point a camera at a sticky trap, and take a photo once a day. Then an algorithm or a human tries to count the insects.

We think this is a poor approach because:

  • Many insects can land on a sticky trap, but still fly away. These will be false negatives. (We know this, through watching hundreds of hours of videos).
  • Camera-based systems have a huge problem with the lenses getting dirty (fields are dusty places). If you have to send someone into the field to clean the lenses, your labour costs will be high.
  • We are getting data from the insects when they are alive and dynamic. Camera-based systems are getting data from dead bugs. With our richer information, we can normally tell the sex and species of the insects.”

How has pest management changed in recent years? What impact does global warming have on the occurrence of harmful insects in crops?

Dr. Eamonn Keogh: "Our sensors provide data that is exploitable at many levels (if and only the grower agrees). For example, all the data we collect is of interest to seed providers, pheromone providers, co-ops, insurance companies, neighboring growers." - Photo: FarmSense
Dr. Eamonn Keogh: “Our sensors provide data that is exploitable at many levels (if and only the grower agrees). For example, all the data we collect is of interest to seed providers, pheromone providers, co-ops, insurance companies, neighboring growers.” – Photo: FarmSense

Global warming has made pest management harder. Insects are critically temperature sensitive, and small changes in temperature means: A) You might see the insects later or earlier than expected. B) You might get three emergences a year, when you used to only have two, C) You might see insects that you normally don’t see.

Can smart pest management and insect monitoring reduce use of pesticides and protect beneficial insects, such as bees? How?

“With ‘dumb’ pest management, you blanket spray and kill everything, including bees (and possibly earthworms).

With ‘smart’ pest management, you may be able to:

  • Spray less, with pinpoint precision in space and time.
  • Use pheromone puffers. This only makes economic sense if you use them on the right days, and at the right time of day.
  • Use biocontrols. Again, this only makes economic sense, if you have good timing information.

So, yes, smart pest management and insect monitoring could greatly reduce the use of expensive and damaging pesticides.”

Also read: FarmSense wins $ 2.2 million in funding for smart pest control

Is pest management without use of any pesticides possible in the future? How can insect monitoring help?

“We strongly believe the answer is yes. It will not happen overnight, but bit-by-bit, crop by crop, we will be able to use less and less pesticides. The European Union has mandated a 50% reduction of pesticides by 2030, they may follow up with further 50% reduction of pesticides by 2040. There is some evidence that a vision for pesticides-free (or at least, drastically reduced) future.”

Can insect monitoring be applied in all crops?

“The technical answer is yes. The economic answer more complex. If you are only making a $ 100 profit per acre, you can’t invest too much in any kind of control or monitoring. Having said that, the price of all technologies falls quickly (including ours). So, many crops today, some new crops next years, some new crops the year after…”

When does it make financial sense for a farmer to invest in insect monitoring? What investment are we talking about?

“The investment can be zero dollars up front. Instead, you pay for a subscription to an insect monitoring service. We have made the barriers to adoption very, very low.”

What are the main (technical) challenges in insect monitoring?

Each Flight Sensor is fully automated, energy independent and can be equipped with basic weather data sensors. Integrated 2G, 3G and LTE connectivity allows reliable automatic data collection. - Photo: FarmSense
Each Flight Sensor is fully automated, energy independent and can be equipped with basic weather data sensors. Integrated 2G, 3G and LTE connectivity allows reliable automatic data collection. – Photo: FarmSense

“It was easy five years ago to get our technology working in the lab. But in the fields you have wind and rain and baking sun, you have dust, you have curious birds, bats, and lizards. So making the sensors robust to all of that was hard (but is not done). Another challenge is doing all this with technologies that are inexpensive. If we build the greatest sensor in the world, but we could never get its price below $ 1,000, we would be wasting our time. So our mantra from day one was: Let’s make it so cheap, that it is not worth stealing!”

What is data mining and why is it so important for insect monitoring and pest management?

“Data mining is the study or how to extract data from large collections of data. Our sensors provide data that is exploitable at many levels (if and only the grower agrees). For example, all the data we collect is of interest to seed providers, pheromone providers, co-ops, insurance companies, neighboring growers.”

What challenges lay ahead when it comes to pest management?

“We think that many of the technical challenges have been solved (of course, you can always be a little more accurate, a little cheaper etc.). Many of the challenges are cultural; growers are often (rightly) conservative when it comes to new technologies. So there is work to be done in education and forceful demonstrations.”

FarmSense plans to launch a FlightSensor model in 2022 that will focus on the navel orangeworm, a pest that is particularly problematic to California nuts.

Join 17,000+ subscribers

Subscribe to our newsletter to stay updated about all the need-to-know content in the agricultural sector, two times a week.

Claver
Hugo Claver Web editor for Future Farming
More about