CMURobotics RI Seminar: Mel Torrie of Autonomous Solutions

As guest speaker for a CMURobotics RI Seminar, titled Lessons Learned Bootstrapping a Robotic Vehicle Company, Mel Torrie of Autonomous Solutions (Petersboro, Utah), describes how he got into robotics in the first place, why he made the jump from academia to a startup, how that startup survived their "near-death experience", what the company has been doing since, and what he's learned along the way. There is a strong agricultural theme, both in his original motivation and in the history and current operation of Autonomous Solutions.

Technology for its own sake vs. benefits

On using robots to make gardening scalable to millions of acres...

You might wonder why I want to turn land management over to robots. Is it because I'm such a geek that I think everything goes better with robots? No, not really. Sure, I think the technology is cool, but I'm not eager to factor human beings altogether out of any activity, not even those that are dull, dirty, and/or dangerous.

I am, however, eager to see the benefits of replacing methods designed to spread a human operator's time as thinly as possible with methods which reintroduce attention to detail to plant cultivation. Granted, that attention would, for the most part, be provided by robotic sensors, processors, and algorithms, but that has an upside as well as a downside.

The downside is mainly that it's unfamiliar and we don't know whether we can trust it. The upside is that robots aren't limited to human senses, but can go as far beyond them as available technology will support, providing information not directly available to us, and they also aren't limited to the time a human operator can afford to invest, provided they are capable of autonomous operation. Potentially this means that instead of a single machine running 40-60 hours per week, you can have many machines running 24/7, vastly increasing the available machine time per land area.

Combine that rich sensory information and machine time availability with manipulators capable of performing basic gardening operations (planting, weeding, pruning, harvesting, saving seed, ...) and you have a platform justifying the effort to develop higher level software which supervises beneficial plant combinations in space (polyculture) and time (crop rotation), which is capable of mixing annuals with perennials (permaculture), which makes room for native plants (particularly those that are threatened or endangered), which also makes room for animals without allowing them to ruin crops, and which can make sure there are enough flowers, throughout the season, to keep pollinators healthy.

Such a system could also learn from experience, even setting up intentional experiments, maintain the genetic diversity of crops, and accelerate the creation of new strains in response to changing conditions. It could also provide exquisitely detailed information in support of operation management.

Technology of this sort would make for a far more varied and interesting landscape (biodiversity) and greater variety of production (diversified income), supporting a more well balanced diet for those dependent upon it. It could also dramatically reduce the energy and other inputs used in crop production, all but eliminate soil erosion and the airborne dust and stream contamination that result, reverse the loss of organic matter in the soil (improving its water absorption and retention and contributing to the sequestration of atmospheric carbon), save many endangered species from the brink of extinction, and bring the practice of agriculture into parity with other aspects of the modern world in its sophistication and the respect it commands, helping to enhance rural culture.

So, yes, that would be very cool technology, but it's really about all of the benefits that could be realized through its widespread deployment.

John Robbins speaking at Stanford

John Robbins speaking at Stanford, before the election, about California's Proposition 37, which would have required labeling of GMO content in foods, and related topics. (Mr. Robbins' talk begins at 3:30.)

an alternative statement of purpose

Cultibotics is about the application of robotics to the cultivation of plants for food, animal feed and forage, for flowers and foliage, for fiber, fuel and feedstocks, for erosion abatement, soil enrichment and landscape enhancement, and for habitat and the preservation of biodiversity. It is also about an approach to robotics that is informed by the best practices of horticulture: minimal tillage, composting, mulch, continuous ground cover, beneficial plant combinations (polyculture), the use of perennials (permaculture), biological methods of pest control, pruning of infested and diseased plant materials, the avoidance of unnecessary soil disturbance or compaction, and so forth. The dream is that robotics can serve as the means by which these best practices can be made to scale to millions of acres.

Re-righting the food pyramid through robotics

There are two main reasons why meat, dairy, and cereal products have come to dominate the dietary landscape. First, humans have a natural taste for animal-sourced foods, a taste we've gradually become more accustomed to indulging as our prowess in hunting increased, a skill which then became less relevant with the domestication of herds and flocks. But secondly, and for the purpose of explaining the current situation, more importantly, the combination of cheap oil and mechanized agriculture has made grain production so inexpensive that the use of grain in animal feed competes with its use in the production of ethanol for fuel.

The other side of that coin is that, so far, mechanization has mainly been applied to grain, soya, and a small selection of other crops that can easily be handled in bulk. Meanwhile, until quite recently, crops like strawberries, tomatoes, and asparagus had to be harvested by hand, making them relatively expensive. Mechanical handling is now becoming available for the more common fruits and vegetables. On the other hand, these crops are also typically heavily dosed with synthetic fertilizers and pesticides.

But mechanization as currently practiced comes with a significant downside. The machines, or at least the tractors that pull them, tend to be large and heavy, compacting the soil over which they pass, which must then be tilled to reloosen it, which kicks up some dust and exposes the soil surface to further erosion, as well as over-aerating the top few inches of soil, resulting in accelerated decay of organic material. The loss of organic material reduces the soil's capacity to absorb and hold water, making production more dependent upon predictable weather, something we can no longer count on.

The solution to all of these problems is to replace conventional mechanical approaches with smart machines, robots, that get their energy from sun and wind, use techniques compatible with no-till polyculture incorporating perennials, and operate autonomously. Using this approach, biological methods could supplant most use of synthetic fertilizers and pesticides, and to the extent these were needed they could be applied sparingly, with precision. This approach would also act to level the pricing of fruit and vegetables as compared with meat, dairy, and grains, making a more balanced diet more affordable.

One additional benefit of the robotic approach is that it could just as easily make room for native flora and fauna, making the same land serve both as native habitat and for crop production. Beyond a certain level of sensory and mechanical sophistication, it becomes a matter of programming to take such factors into account.

Reviewing “Farmerbots: a new industrial revolution” by James Mitchell Crow

Writing in issue 2888 of New Scientist, James Mitchell Crow introduces us to the notion that robots will, sooner or later, be tending the crops we depend upon for food, and takes us on a whirlwind world tour of some of the people working to bring this about and some of the technologies that have already been developed.

He begins with Simon Blackmore, of Harper Adams University College, who tells us about robotic technologies that have already found their way into new tractors, implements, and combine harvesters. Blackmore also discusses the energetics of cultivation, saying “Why do we plough? Mainly to repair the damage that we have caused with big tractors. Up to 80 per cent of the energy going into cultivation is there to repair this damage.” He proposes an altogether different approach, using light-weight, autonomous machines. Crow summarizes the requirements list for these machines thusly: “These agribots need to have three key abilities: to navigate, to interpret the scene in front of them, and to be able to help the farmer, by blasting a weed, applying a chemical or harvesting the crop.”

Addressing the first of these requirements, navigation, Crow moves on to Germany, to mention Arno Ruckelshausen of the University of Applied Sciences in Osnabrück, who is developing RTK-GPS, a geolocation technology with a resolution of 2 cm, for BoniRob, a four-wheeled rover that also uses spectral imaging to distinguish green plants from brown soil, and which remembers the location of individual plants and returns repeatedly to each to monitor their growth. Ruckelshausen also intends to equip BoniRob with a precision chemical application system based on ink-jet printer technology, which would apply microdots of pesticide directly to the leaves of weeds. For those averse to using herbiicides in any amounts, Crow mentions flame guns and lasers as alternative methods of weed control. He also goes into the use of a similar approach for fertilizer, using sensors to gauge how much is needed by individual plants and supplying only that much.

The next stopping off point is Australia, where Salah Sukkarieh, of the Australian Centre for Field Robotics, makes the point that funding is relatively abundant for mining and defense projects, but relatively paltry for agricultural robotics, meaning that difficult problems like machine vision will have to be worked out first for other applications first and then reapplied to agriculture. Evenso, he predicts machine vision systems will be available in approximately three years.

Crow makes a quick side-trip back to Europe to mention the HortiBot, another four-wheeled field rover that already uses machine vision to identify weeds and apply chemicals to them.

Then on to Japan, a country with an aging population that currently only produces 40% of the food it consumes. Despite the shrinking labor pool, the government of Japan hopes to reduce their dependency on foreign imports by increasing the amount grown domestically to 50%. To accomplish this, the Ministry of Agriculture, Forestry and Fisheries has tapped Noboru Noguchi of Hokkaido University to lead a 5-year, $8 million effort to automate all aspects of the cultivation and harvesting of rice, wheat, and soya, and to bring this machinery to market by 2014.

Continuing with the theme of a shrinking labor pool, Crow quotes Eldert (E.J.) van Henten of Wageningen University, The Netherlands, as saying “Work in agriculture is not interesting, prestigious or usually very well paid. It is physically demanding and dirty. People prefer to go to the cities and work in factories or in office jobs. While the population is growing and needs to be fed, a rapidly shrinking number of people are willing to work in agriculture.”

I presume Henten was referring to hired farm laborers rather than to farmers, but even farmers may soon be in short supply. BBC Radio's Farming Today has recently conducted several interviews with people involved in various aspects of agriculture, in which the interviewees have expressed concern that they don't see a new generation of farmers ready to take the places of their elders.

Crow next dives into the economics, beginning with the California raisin industry, which, after being squeezed by a price-crash following a bumper harvest in 2000, began adopting a mechanical harvester adapted from a machine used to harvest ripe grapes for wine. By 2007 almost half of California's raisins were being harvested mechanically. Generally speaking, the use of robots can reduce costs for field operations such as weeding, even after factoring in the cost of machinery and maintenance. One Danish study of organic farming concluded that agribots could reduce the cost of weeding by half.

This article strikes me as being reasonably well-researched, if perhaps a bit simplistic. I see some familiar names. On the other hand, there is no mention of Integrated Pest Management, in which biological control methods are used to reduce the need for chemical pesticides, and for the practice of which autonomous machines could certainly provide assistance. There is also no mention of the significant advantage to be gained, for example in crop breeding, from the generation of detailed, plant-by-plant growth records, nor of the potential for making use of robots to replace monoculture with intense polycultures, including mixtures of deep-rooted perennials and annuals. But comprehensive or not, it's good to see such an article appearing in a publication like New Scientist Magazine, and I'm grateful to Mr. Crow for having gone to the trouble.

Robohub: news, views, and everything robotic

Robohub is a new robotics website with a deep history. Brought to you by the same group that produces Robots Podcast, Robohub is, at the outset, among the best connected sources of robotics news, and also offers interviews with and editorials written by robotics experts, microlectures on tightly focused subjects, and tutorials that will help you get your own projects up and running. Robohub is your window into the world of robotics.