Israeli agriculture integrates latest digital technology to increase efficiency.
By ZVI HELLMANagriculture and technology cartoon 521(photo credit: AVI KATZ)
At first glance it looks like any other gathering of high-tech investors and eager young entrepreneurs looking for investment.But listen again and it becomes clear that the talk about improving yields may actually be about agricultural crops. The references to herd behavior, are about cattle, not nervous market investors. At the Agrivest Summit, a conference on agricultural investment sponsored by the Ministry of Industry, Trade and Labor in Israel in May, “seed funding” takes on new meaning.The buzz here is that the world is on the cusp of another agricultural revolution.The 1950s and 1960s brought the Green Revolution, improving crop varieties using scientific breeding. The next leap forward, according to industry leaders, will bring the extraordinary advances attained over the past two decades in electronics, telecommunication, data analysis and biology labs to fields, pastures and fisheries.Computers will monitor every aspect of a plant or animal’s life cycle to ensure optimal yields. Robots will plow fields in place of manned tractors. Genetic engineering will create new plants resistant to drought conditions and pesticides. Fuel for cars and jets will be harvested from genetically modified algae. These are not science fiction stories. They are just a sampling of research and development efforts in new agricultural enterprises.Fittingly, the conference was held in buildings constructed over a century ago in Mikveh Israel, the first modern agricultural school in Israel, established in 1870 by pioneers infused with a vision of a future Jewish state based on an agricultural society.Twenty-first century Israel is decidedly not agricultural. It has become a hyperurbanized country in which youths dream of office jobs in technology parks, not picking oranges or harvesting olives.But in one sense an early Zionist vision has been realized: Israel has become a world leader in agricultural innovation, mostly out of necessity, given a relative paucity of arable land, extremely limited water supply, a small agricultural labor force and significant distance from the world’s major markets. From drip irrigation through the development of new strains of potatoes and tomatoes and methods for raising fish in the desert, Israeli agriculture has consistently contributed new ideas to the world.Scientists and mathematicians Combining Israel’s high-tech know-how with its agricultural innovativeness holds the promise of surprising synergies. One of the most striking aspects of the Agrivest conference was the sheer number of computer scientists and mathematicians attending the sort of gathering that only a short time ago would have been of interest only to farmers.
IBM, which has an agriculture analytics division in its Haifa Research Lab working with about 3,000 IBM employees worldwide focused on agricultural optimization, sent two representatives to Agrivest.“Agriculture is fundamentally a mathematics problem,” says Richard Kottmeyer, an agriculture and food production leader at IBM Global Services in St. Louis, and a guest speaker at the conference. Kottmeyer obviously enjoys the look on people’s faces when he tells them that his job at IBM is focused on agriculture. He believes that future progress in agriculture will depend on getting mathematicians on board.“In agriculture, there are many variables and unknowns in a complicated formula for getting optimal yields,” he explains. “We are fundamentally a mathematics formulas company. We are good at working with new technologies and bringing them to market.Who could be better than IBM, then, for the challenges of agriculture?” The challenges are indeed daunting. The global population stood at just 2.5 billion in 1950. It is now over 7.2 billion and on track to reach 9 billion by 2050. At the same time, the caloric intake per person is steadily increasing, meaning that the world will need a 70 percent increase in caloric production relative to the year 2000 by 2050, at which point an astounding 10.2 quadrillion calories per year will have to be produced.At the same time, annual yield improvement has been falling in the last 20 years. Water use has been growing at twice the rate of population increase, while the amount of arable land will not change significantly in the foreseeable future and oceans have been tapped for fish nearly to capacity.There is tremendous scope for improvement in efficiency. For example, it takes 7 kilos of animal feed to produce 1 kilo of beef. The comparable rate for chicken is 2.5 to 1, with fish representing the most efficient conversion rate of 1.5 kilos of feed to 1 kilo of end product.Significant yield improvement, according to speakers at Agrivest, can now be attained by transferring the same tools used in the technology world to the world’s farms.They speak of “precision agriculture” using sophisticated electronics and sampling and statistical tools. Forget the romantic image of the scythe and plow as the farmer’s main implements and replace them instead with smartphones, GPS, and remote sensing devices.From the perspective of the investors, now is the time to get in on the early stages of the agriculture-tech game. Only 3 percent of the private equity investment is currently channeled to agriculture, compared to 8 percent in telecoms, 12 percent in health enterprises and 14 percent in information technology. Governments are also awakening to the opportunities and needs in agricultural technology. The Israeli Ministry of Industry, Trade and Labor, for example, has established an agricultural technology incubator program.Ilan Arbel freely admits that he had never seen a cow up close when he had the idea for CattleSense, a start-up he founded that enables ranchers to keep track of the health conditions of thousands of cows scattered over immense grazing territories just by glancing at their smartphone screens.Arbel, a retired air force major with degrees in physics and industrial engineering and a résumé that includes extensive experience as a senior business development executive at several computer software firms, tells The Report that his interest in the cattle industry was piqued a few years ago when he stumbled upon a report that the amount of beef produced in the world had remained flat for over a decade.“Between 1991 and 2001,” says Arbel, “the number of heads of cattle brought to slaughterhouses remained steady at 93 million to 94 million per annum, despite rising demand, which in turn led to increasing consumer prices.”Turning over those numbers in his mind, Arbel decided that there was a business opportunity, if only he could figure out why this was happening and find a solution. “I am a mathematician by training,” he says. “This was a puzzle that I had to solve. It turns out that cattle ranchers simply could not increase the number of weaned calves reaching adulthood because they could not keep track of them, and they were consequently dying due to diseases and other natural causes of attrition.”In countries with vast rangelands, such as Australia, Brazil, Argentina and the United States, cattle growers can be responsible for herds with as many as 10,000 heads of cattle freely grazing over immense territories.“Imagine a rancher with that many cows roaming over an area equal to the western Galilee,” explains Arbel. “He may be in direct contact with the animals only once a year. This makes it impossible physically to keep track of where they are or of their condition. As a result, for every 1,000 calves born, only 500 survive their first year.”RanchersCattleSense intends to change that by giving ranchers all the information they need to keep track of their herds and make informed daily management decisions on an ongoing basis, at their fingertips, on their smartphones. If a disease is spreading through a herd, or the cows are grazing in inferior pastureland and losing weight, or are overheating, ranchers using CattleSense can pinpoint exactly where the problem is located and get directly to the location to treat the herd. The system can also track which bull has mated with which cow, how successful the mating efforts have been in terms of resulting pregnant cows and provide immediate statistics on which cows are in heat and likely to be receptive to bulls, so that ranchers can direct the bulls to those locations most efficiently.Not surprisingly, this requires a significant amount of mathematics and statistics to analyze a large amount of data. “We constructed a network of sensors, with the sensors located on collars on the cows,” says Arbel. “The information gathered by the sensors is uploaded by satellite link and is available to ranchers on their computers and smartphones, along with GPS maps showing where each cow is located alongside the animal’s current state of health. Using this system, the number of calves surviving to adulthood jumps on average from 500 out of 1,000 to 800. That is a dramatic increase.”The CattleSense collars are non-invasive, making use of new technology developed by scientists on its staff to measure cow health entirely externally. The sensors work on solar energy soaked up while the cows wander through the pastures – a necessity, since there is no way to plug the collars into electrical sockets in the open fields. In another innovation, CattleSense has developed a low-cost radio frequency network connecting each group of 500 sensor collars together, a sort of Facebook for cows. The data uplink is conducted at only a few points in the total network instead of from each animal, resulting in large savings in costs.AutoAgronom, a Yokneam-based three-year-old start-up, aims to give plant growers what CattleSense gives ranchers, the data they need for optimal yields, along with an automated system deciding the optimal amounts of water, fertilizer and oxygen to feed crops. Noam Ilovich, VP of business development at AutoAgronom, calls his company’s approach “listening to the roots.”The sensors are attached to plant roots and form a network monitoring the condition of all the plants in a field.“Farmers always face the challenge of determining how much fertilizer to use, how much water, and so forth,” explains Ilovich. “They’ll use generic statistics tables, experience, their ‘green thumbs.’ With our sensors and software, they can use the roots themselves as the source of expertise.We detect chemical and physical values to automatically maintain optimal water, oxygen and nutrient levels. Our algorithms combine machine learning with monitoring.”Ilovich is a Technion-trained software engineer who has worked at Elbit and Microsoft and founded start-ups of his own. He was so impressed when he started working at AutoAgronom, he decided to invest his own money in it. He says the leap from computers to plants was not so hard.“I have plenty of experience working with customers. I know the business world. It wasn’t a difficult move,” he says. “What is a challenge is that we need to explain to farmers why they need the technology in a way that we don’t need to explain in software. It takes time.”AutoAgronom reports that with its system, farmers can use 50 percent less water and 70 percent less fertilizer while boosting yields by 15% – very impressive numbers. “There are nights I don’t sleep because the work I am doing is so exciting and important,” says Ilovich. “I didn’t feel this way in any job I’ve had before. It is unbelievable. We are doing something real and significant.”
