PHOTO: PETER ROEK


GLOBAL VIEW ▶▶▶


What is your preference: drones or satellites?


I BY MATT MCINTOSH


n the race for more eco- nomically efficient and practical agronomic data, do drones or satellites provide you with the larger ben-


efit? It seems to be one of the most critical factors when it comes to investments in tech and agronomy, but most articles about the subject ignore this question. Companies offer- ing satellite mapping services tout the fre- quency of flyover and high resolution as THE difference makers in terms of making practi- cal, real-time decisions on the farm. This makes sense to a degree if I was managing thousands of acres. But if I’m farming 300 acres, walking every row is much more feasi- ble, and it is likely preferable. In such a case, the occasional drone flight might be the tick- et. Indeed, it seems somewhat obvious, and I do not wonder whether this is the reason drone services appear more popular than sat- ellites in Ontario; whether this is true or not I am unsure – it is just a personal observation from a small corner of the globe.


Satellite imagery, drones, or maybe none? Know what you are looking for and how you can use it, then go find it.


Establishing goals before investing But for argument’s sake, let’s assume the obser- vation is true. Why, then, does one encounter arguments saying one technology will ulti- mately overtake the other as universally impor- tant? Like everything in agriculture, the scale of analysis afforded by drones and satellites seems to reflect the diverse needs of diverse farm operations. Fundamentally, is it not more important to know what goals you are trying to accomplish before investing in either – or in- deed any – new technology? I would not buy a


massive Case 9240 if the older 1660 will do the job; how is mapping technology any different? Of course, whether one can even use the data provided by satellites and drones is another story. But I would argue this also comes down to first establishing specific goals, then looking for the right tech to match. I recognise service companies are trying to sell a product, so a lit- tle extra hype is to be expected. However, I wonder whether greater emphasis on helping individual farmers develop and act on goals should not play a larger role.


Self-learning algorithm gets better at weed detection


B 46 BY LEO THOLHUIJSEN


y employing so- called deep-learning systems for crop recognition, ma- chines that combat weeds site-specif- ically with a minimum of labour and/or chemi- cals are getting ever closer. It works like this: you start with a blank algorithm that knows nothing and train it by using examples. You show it an image of a beetroot and say, ‘this is a beetroot’. If you then proceed to show it a


potato plant without any further instructions, the algorithm will think it is a beetroot. So you tell the robot that it is in fact a potato plant. By doing this, you can refine the algorithm by con- tinuously showing it new images. After showing it (tens of) thousands of exam- ples, the algorithm learns to detect specific as- pects of crops. Which characteristics it specifi- cally considers is unknown; call that the large black box. That this is not easy was recently confirmed in a field test. At first, the system worked exception-


▶ FUTURE FARMING | 1 november 2019


ally well. All potatoes were sprayed, beetroots were nearly all left alone. A week later, things did no go as planned. The beetroots had grown, and the potato plants were already nearly dead. The algorithm never encountered such large beet- roots or sprayed potato plants before. Several beetroots were sprayed, and several potato plants were not. This problem can be easily solved by collecting new images. The system can be retrained. But it is a big challenge if we also want to detect crops and many different weed varieties in the future.


Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24  |  Page 25  |  Page 26  |  Page 27  |  Page 28  |  Page 29  |  Page 30  |  Page 31  |  Page 32  |  Page 33  |  Page 34  |  Page 35  |  Page 36  |  Page 37  |  Page 38  |  Page 39  |  Page 40  |  Page 41  |  Page 42  |  Page 43  |  Page 44  |  Page 45  |  Page 46  |  Page 47  |  Page 48  |  Page 49  |  Page 50  |  Page 51  |  Page 52