Friday, December 20, 2013

FRE2014 to be held in conjunction with DLG Field Days

Hosted at least in part by the University of Hohenheim, the 2014 Field Robot Event will take place in conjunction with DLG Field Days (brochure, fact sheet), June 17-19, in Strenzfeld, just outside of Bernburg, Germany.

Saturday, November 30, 2013

regarding the transformational application of robotics to agriculture

Robohub's focus series on agricultural robotics either is nearing or has arrived at completion, however the work of applying robotics to agriculture has barely begun. My own contribution to that series turns on the idea that robotics is a fundamental, revolutionary development, with the potential to transform everything it touches, and, by implication, that roboticists should embrace that potential and approach their work as an opportunity to change the world for the better, both generally and in the context of agriculture.

Sunday, November 03, 2013

elaboration of "keeping dust off optics" (2009Mar01)

Realizing that I really should go to the trouble of at least diagraming what I am about to describe, as I have in the past I will wince and proceed, preferring to get the idea out rather than wait on a diagram that I may never get around to creating. As a stop-gap, I will end with links to several relevant Wikipedia articles that do include diagrams and/or images.

In the original post on this topic, I only described the air flow, without going into detail about filtration and drying of the air used. I intend to amend those omissions here.

Since the needed volumetric supply rate of air is fairly small, significant compression (enough to ensure no condensation on the optical elements) should not represent an onerous power requirement. However, even simple impeller turbines can be worn by dust particles, and filters last longer if coarse material is removed before reaching them, so the first stage should be a cyclonic separator.

As the input to the compressor, a slight vacuum will be pulled on the air as it passes through this first stage. If that air is at or very near dew point, condensation may occur inside the separator. (If condensation does occur, most of the heat removed from the condensing water vapor will remain with the air, raising its dew point.) Packing the center of the separator with a downward pointing cone of stainless steel wool will provide that condensation with a surface to adhere to and wick down, collecting in drops at the point of the cone. The moist stainless steel wool will also help remove finer dust particles that would otherwise pass through the cyclonic separator. (The finer the stainless steel wool, the more effective it is likely to be, but also the more likely that it will need an occasional back-flush with steam.)

Next comes the compressor. This should be a very small impeller turning very fast, to achieve significant compression with a low rate of flow. The chamber enclosing the impeller may also act as a second-stage cyclonic separator. Any condensation remaining in the air flowing into the compressor should be flung against the outer wall of that enclosing chamber or reevaporated instantly.

Following the compressor is a HEPA filter, to remove any remaining dust. The clean air from the HEPA filter is then directed into a box containing the optics (lenses and image sensors). That box will need to be strong enough to withstand mild pressurization, and should be sealed except as described below, although the pressurization will help ensure that any imperfections in the seal don't result in dust infiltration.

Rather than exposing the optics directly to the outside air, they should be protected behind a lens cover in the form of a wafer of flat, very clear glass or crystal. That wafer should be etched on its outer edge with widely-spaced grooves, perpendicular to the flat sides, to serve as air channels, and on it its outward-facing side with groves which begin at the groves in the edge, tapering to nothing as they spiral toward the center at a shallow angle. By mounting such a wafer in such a way that air passing out through the groves around the outer edge is forced to turn and flow in an inward spiral across the face of the glass, an air-flow, reminiscent of that in a hurricane, will be set up, with filtered air spiraling inward next to the lens cover, pushing out a bit, and then spiraling back outward a short distance in front of the lens cover, keeping the lens cover itself free of dust and small droplets in the ambient air.

Now for the promised Wikipedia links…

Friday, October 18, 2013

Robohub focus on agricultural robotics has just opened a new focus on agricultural robotics. Robohub focuses last about a month, so bookmark this page and check back often. There will be a lot more than is there at the moment!

Wednesday, October 02, 2013

Saturday, September 14, 2013

real potential vs. hype

Whenever one recognizes unrealized potential in some new development, there is a temptation to attempt to bring it into being by overselling it, glossing over the difficulties of making that potential real and promising greater benefit and less collateral damage than is actually likely to result.

This is an even more poignant issue in robotics than in other fields, as robots have long been a favorite, abundant source of material for authors and screenwriters, who invest them with whatever qualities and capabilities serve their purposes, frequently without giving serious thought to the engineering effort which would be required to produce these attributes. Consequently, the general public is somewhat confused about and unimpressed by the current state of the art.

Since starting this blog, just over seven years ago, I've tried to make it clear that I've been talking about technology that is clearly within reach but, for the most part, not yet in existence. Over that seven years there has been significant progress, in the technology, and even more so in my own cognizance of the current state of that technology and of ongoing research and development work.

Even so, I still find myself faced with the conundrum that by pointing with such certainty to a potential that can only be fully realized through an iterative development process founded on intensive collaboration between two groups of people who as yet hardly even talk to each other (roboticists and horticulturists) I undermine my own credibility and that of the vision I strive to share.

Nevertheless I am certain, more certain that ever, that machines capable of performing the full range of gardening tasks autonomously can be developed, and that a combination of automated factories and economies of scale can make them competitively affordable, as compared with the combined costs of continuing practices which are currently so dominant as to seem anointed and unassailable.

Above all I am certain that this approach to crop production and land management can be the essential ingredient enabling solutions to a whole set of intractable problems, from malnutrition and (lack of) food security to rural poverty to dwindling biological diversity to spreading deserts to contamination of air, streams, and oceans.

It is unfortunate that many of those who might read this would see it as hype, for it is nothing of the sort. It is simply the recognition of a crucial potential inherent in the development of robotics, and the attempt to infect others with that recognition, rendered imperative by the gravity of its implications.

Therefore I persevere.

Thursday, August 22, 2013

a term worth bringing back: "husbandry"

Perhaps its similarity to the word "husband" has contributed to its decline, as economics has forced women into the workforce in droves, including many who would have been happier as homemakers, and marriages have been recast as partnerships between undifferentiated equals.

Whatever you may think of all that, in the context of land management, there is no other English word which is quite so perfectly nuanced as "husbandry", since it implies as much attention to the care and conservation of resources as to their production and use.

Friday, August 09, 2013

if it'll happen eventually anyway, why worry?

I am reasonably confident that something resembling my vision for the application of robotics to agriculture will eventually happen, so why do I even bother trying to call attention to it?

The question isn't whether it will happen, but how soon and, more importantly, how much further damage will be done before it does finally happen?

How soon is largely a function of resources, both public and private, for R&D. Public funding is somewhat problematic at the moment, of course, with the government in danger of being shut down through the inept stubbornness of certain congress-critters. Assuming that doesn't happen there may be a little money available here and there, through ongoing programs, but we're not talking about a determined effort such as that which put men on the moon. Appropriate as it might be, that sort of approach just isn't in the cards for now.

On the other hand, private money could be more than enough to get things moving. No doubt investors are a little gun shy, given the legal battles still underway in the smartphone industry, and leery of tying up their money in the hope of a payoff that might not come for a decade, maybe even two. The first concern could be eased somewhat by the establishment of a FRAND-based IP consortium. The worry over time-to-payoff could be addressed by identifying development milestones that would be marketable in specific circumstances, as Harvest Automation has done, beginning to generate a revenue stream while still deeply engaged in development.

The question of additional damage is both more looming and more difficult to wrap one's mind around. The variety and scale of environmental insults being perpetrated by conventional agriculture are staggering, and yet this is business as usual. Even if every element of a viable alternative were available today, it would still take time to convince many farmers of the need to change, time to replace equipment, and time to restore soil fertility. Meanwhile those with the most to lose, the chemical companies whose product portfolios read like a litany of sins, are likely to fight change tooth and nail, clinging to their poisonous, monopolistic business models as long as they possibly can.

So there is some cause for worry that by the time everyone is convinced of the need for change, there'll be precious little left to save, and we'll find ourselves surviving on tube-grown GM-algae and vat-grown meat, while weeds that have developed resistance to all of our poisons take over field after field from crops with too little remaining genetic diversity to cope with climate change in addition to everything else being thrown at them.

We need to make use of the best practices we can manage right now, given the small percentage of people engaged in agricultural production and type of equipment currently in their hands, and bring even better practices to bear as willing hands and robotic technologies make them possible in urban environments and on an agricultural scale. That is the path to saving all that can still be saved.

Saturday, July 27, 2013

“Don't worry, we know what we're doing.”

Frankly, this assurance, the claim on the part of "experts" (those who are in the business) that they know what they're doing, is more often false than true, at least insofar as the implication is that they know all there is to know about the consequences of their beliefs, choices, and actions.

Which "experts" you ask. What business? It's an observation about people and the nature of knowledge, so it hardly matters, except as the stakes may be higher in some cases than in others. One such case is the set of practices collectively known as modern agriculture, which includes the use of GMOs, pesticides, herbicides, and concentrated fertilizers, and which, because it is applied over such a large area, can fairly be characterized as an experiment being conducted on the planet – the only planet we have.

In an article published on the Scientific American website, Professor Nina Fedoroff of the King Abdullah University of Science and Technology and Pennsylvania State University, states the following:

Most early alarms about new technologies fade away as research accumulates without turning up evidence of deleterious effects. This should be happening now because scientists have amassed more than three decades of research on GM biosafety, none of which has surfaced credible evidence that modifying plants by molecular techniques is dangerous.

On this point, Professor Fedoroff is clearly speaking from expertise, but notice that the issue she addresses is whether the process of modifying plants by molecular techniques is inherently dangerous, not whether specific modifications might be so. In the same paragraph, Fedoroff goes on to fume somewhat recklessly:

One scare story based on a bogus study suggesting a bad effect of eating GMOs readily trumps myriad studies that show that GM foods are just like non-GM foods.

Further on she states:

Herbicide-tolerant crops have made a major contribution to decreasing topsoil loss by facilitating no-till farming. This farming method reduces CO2 emissions from plowing and improves soil quality.

At this point I must express gratitude, for her having mentioned the advantages of no-till farming, but there are better ways to manage it than through the application of herbicides. There's the matter of reduction in yields when herbicides are applied and the plants must invest energy in resisting them. Also one must wonder how much attention has been given to the question of whether the genetic modification which provides plants with resistance might itself be problematic, perhaps resulting in byproducts which are toxic to animals and humans. Returning for a moment to an earlier passage, Fedoroff claims:

We can now use these methods to make precise improvements by adding just a gene (or two or a few) that codes for proteins whose function we know with precision.

This is blatant conceit. The proteins genes code for commonly have more than a single function in the development and/or physiology of an organism, and may also have collateral effects that wouldn't be termed "functions". To say that one knows the function of a protein "with precision" is to claim to to know all potential effects of that protein under all conceivable conditions. In her conclusion Fedoroff displays almost childlike naiveté regarding corporate veracity and commerce:

If the popular mythology about farmer suicides, tumors and toxicity had an ounce of truth to it, these companies would long since have gone out of business. Instead, they’re taking more market share every year. There's a mismatch between mythology and reality.

I do not suspect Professor Fedoroff of being a complicit apologist for industry, but I do suspect her of habitually misplaced trust.

Reposted from Lacy Ice + Heat.

Sunday, July 07, 2013

can robotics help make farming a less dangerous occupation?

If your notion of what robots are like comes from movies and television, or reports about drone strikes, you might think I'm daft for even suggesting that the application of robotics to agriculture might make it a safer occupation.

Well, granting that some of those fictional portrayals are poignant warnings about particular potential developments in robotics, they aren't representative of the field as a whole. Most robots currently in existence are either the industrial variety, isolated from people because they aren't designed to recognize our presence, or vacuum sweepers – too small and light to do any serious damage to an adult, so, provided you don't allow an infant to crawl on the floor one of them is cleaning, it makes little difference if they can't tell a person from a bookcase.

On the farm, robotic technologies have been finding their way into tractors and other large equipment, but mostly not yet the sorts of technologies that would allow such equipment to identify a human in their path. So, at this stage, to the extent that these machines are left to drive themselves, one might argue that they are actually less safe than equivalent machines under the direct control of human operators. Perhaps, perhaps not. The straight lines, smooth curves, and complete coverage with minimal overlap those automatic guidance systems produce require considerable concentration when performed by a human operator, not leaving much spare attention for the path ahead, and continuous concentration tends to make the mind wander. This is why you'll occasionally see human-operated equipment lose the pattern and go rambling across a field at odd angles, until the driver snaps to and gets it back on track.

A more interesting question than whether robotic technology currently in use results in greater or lesser safety for a bystander is whether there might be a point, in the improvement of such technology, beyond which one is safer standing in the path of a tractor than riding in the cab. Tractors are rather easy to overturn; they sometimes catch fire; they occasionally get hit by lightning; they rock, lurch, buck, vibrate, and produce a mind-numbing noise. Hydraulic fluid spraying out of a crack can cut through flesh. Unguarded PTO shafts can grab loose clothing and pull the wearer to their death. Tractors are built for raw torque and force, and possess a potential for mayhem comparable to a bulldozer, in the event their operators should pass out or otherwise become incapacitated.

Automated guidance systems can sound an audible alert as they are nearing the end of a row, and shut down if the operator doesn't respond, but there's a lot more that can be gained from the utilization of robotic technologies. A whole range of sensors is available, from ordinary digital video to infrared to LIDAR, and a good deal of work has already gone into integrating the information provided by multiple sensors. Even better, these technologies are moving towards being reasonably priced off-the-shelf plug-ins, as with the Kinect. Millimeter radar, when it becomes widely available, will add another, extraordinarily useful option.

Tractors and other such large equipment present a special challenge in that, in operation, they are nearly always surrounded by a cloud of dust. Moreover, the nature of that dust varies depending on the type of rock form which the mineral portion of the soil weathered, or whether the source of the dust is soil disturbance (tillage) or rough handling of dry plant material (harvesting). A technology that effectively penetrates one sort of dust may not work well when confronted with another. Stereo sensors and processing which extracts useful information from noisy data will undoubtedly be important components of any thoroughly adequate solution.

So much for adapting robotic technologies to conventional equipment. In my opinion the greatest gains in safety are to be had from reducing the size and power of the equipment in use, refocusing on detailed operations like planting single seeds, pruning single twigs and leaves, picking pests one by one, and so forth. Such equipment would be inherently safer, since it simply wouldn't possess the capacity for wholesale destruction. There would still be issues, of course. A manipulator that can snip off a branch could do the same to a finger, and almost anything can be dangerous when near the eyes, but the scale of the threat imposed by smaller machines would be at least an order of magnitude lower, based on their capacity for mechanical force alone.

Additionally, these machines would be making fine-grained decisions, instead of simply running until directed to do otherwise. Each detailed operation would first need to pass a test as to whether it were a good use of the machine's time and available energy, then the motions comprising each operation would be crafted for efficiency and minimal disturbance of the surrounding environment, foliage included. If you're sensing and planning paths around branches and leaves, something as solid as a human will appear as inviolable as a brick wall. Layer on detection and avoidance of warm-blooded animals or vertebrates, and the result should be a very reliable safety regime.

There are other hazards around a farm besides the equipment, but robotics can, and sooner or later will, all but eliminate threats from that source.

Thursday, July 04, 2013

why green activists should support and help shape robotics

Call it what you like, organic farming, regenerative agriculture, agroecology, or whatever else, unless all you're talking about is green window dressing on otherwise conventional techniques, involving routine tillage and unrelieved stands of uniform annual crops, your going to find that most of the available equipment isn't well suited to the approach you're proposing.

Take permaculture for one example. Conventional agriculture knows something about permaculture, of course. There's orchards and vineyards, and a few crops, like asparagus, that regenerate annually from rhizomes. Plant the trees in rows, far enough apart that you can fit a small tractor between them, and you're in business. For perennial crops that are mown down annually, the row spacing can be half the track width of the equipment you're using, since that equipment can straddle the rows. This isn't optimal; the space between the rows generally goes unused. But this approach allows the use of familiar equipment.

On the other hand, if the crop you have in mind to grow is a perennial version of wheat (currently in development), and you're competing in a commodity market with others growing wheat in the usual manner, as an annual, that space lost to tracks for equipment wheels starts to be significant, cutting into your available acreage and, because margins for commodities are tight, profitability.

For another example, let's consider polyculture. One such that most people have at least heard about is the traditional mixture of maize, beans, and squash in the American southwest. Think about how you would handle this cropping system mechanically. Can all be planted at once, or will planting require multiple passes. Can you harvest the beans without damaging the maize or the squash? What about weeding?

Now consider a system incorporating both permaculture and polyculture, say growing your maize, beans, and squash in the spaces between apple trees. How will you get your equipment onto the land to get anything done at all? Perhaps make sure the wheels of your tractor and other equipment always follow the same tracks, but that's an awkward solution. A better solution would be to install elevated rails (expensive), or posts which could be used to support relocatable rails, or pads on which legged machines could place their weight.

The last of these approaches obviously requires robotic equipment, with articulated legs. The posts and movable rails approach also requires at least a robotic arm, to handle the rails. Permanent elevated rail could be used to suspend anything, but if you've gone to the expense of installing it you'll want the other advantages robotic equipment can offer.

Those other advantages derive from the potential of robotics to make the best practices of gardening scalable – practices an experienced gardener might apply on a quarter acre of land but not even attempt on a five acre plot, for lack of time. I've mentioned many of these at one time or another, but in a nutshell the option to make use of them is a function of how many eye/brain/hand-hours you can bring to bear per area per growing season.

Autonomous operation allows one person to delegate this work to multiple/many machines, which (assuming they aren't using energy-intensive techniques like plowing), can operate 24/7 if need be. Sensors substitute for eyes, processors for brains, and mechanical arms and manipulators for hands. Given these basics, robotic machinery can, for instance, pick slugs and larger insects instead of poisoning them, prune diseased/infested plant parts instead of spraying, and deal just as easily with a mixed stand including perennials as with a uniform crop.

To be blunt about it, if you want to reform agriculture, you're going to need a lot of help from robots, and the sooner you realize this and begin to participate in shaping the future of robotics, the better the outcome will be.

Saturday, June 29, 2013

FRE tasks becoming more interesting

A sure sign that the state of some art is improving is when the difficulty of the tasks that must be performed in a competition is increased. That was the case with this year's Field Robot Event in Prague, which ended earlier today.

There was the usual navigating through rows of plants, but this year's tasks included detection of weeds (simulated by maize and oil seed rape) among the sunflowers, and detection of diseased or otherwise damaged sunflower plants.

In addition to this, team members following the robots were not allowed to carry any electronics that might be used to provide them with instructions. They could manhandle the machines to help them reorient at the ends of rows, at the cost of a penalty, but aside from that the machines were required to complete each task autonomously.

In my humble opinion, these changes are a hopeful sign that the intention is to move the competition towards real world applicability.

Saturday, June 22, 2013

sentry duty for ‘cultibots’

Earlier this month, at TED Global, science fiction author Daniel Suarez made a powerful argument for an international legal framework prohibiting the development of lethal autonomous systems, otherwise known as killer robots. He went on to suggest that machines designed to recognize menacing rogue behavior in other machines and raise an alarm could be used to protect against any killer robots that might be set loose despite the ban.

Perhaps that task, detecting the presence of a killer robot and raising an effective alarm, could be distributed across a wide variety of machines, particularly those working outside. ‘Cultibots’ would be ideal for this purpose, as they would also be geographically distributed, providing better coverage. I wouldn't suggest requiring such functionality, at least not yet, as that would simply delay deployment of robots for other purposes while this relatively sophisticated capability was developed, but once it becomes available it should be incorporated into any machine, with sufficient capacity to support it, that might be in a position to make early contact with a rogue.

Another, even more sophisticated capability that should be developed and incorporated into all types of robots is the ability to recognize and block any attempt to coopt or destructively repurpose themselves, whether on the part of a nearby machine or through a remote connection.

Suarez went on to suggest that machines dedicated to dealing with rogue robots might snare them and haul them away, but he says that, despite that they're dealing with other machines rather than people, they shouldn't be allowed to autonomously decide to destroy those other machines; that decision should always be made by a human being. Under most circumstances, other types of robots should avoid engaging a rogue, if possible.

A situation in which it might not be possible to avoid engagement would be if a human being were at risk, as expressed by Isaac Asimov's First Law of Robotics: “A robot may not injure a human being or, through inaction, allow a human being to come to harm.” Considering that an attack on a human might be the first clue that a robot was a rogue, and that a second's hesitation might make the difference between life and death, any robot in the vicinity should be ready to do what it can to block such an attack, perhaps even suspending other duties so as to be able to pay close attention whenever both a human and a robot of unknown trustworthiness are nearby – devoting any spare processor cycles to running scenarios and developing contingency action plans.

Of course, governments will want the ability to remotely repurpose whatever robots are available, not only to deal with rogues but to deal with all kinds of emergencies. However, there should be no way for them to turn ostensibly innocuous machines into combatants (beyond the requirements of the First Law), much less into autonomous killers. To ensure this, remote repurposing of ordinary robots should be constrained to a predetermined, short, fixed list of alternative modes. Any changes to that list should require either a ‘brain’ replacement or the physical presence of a factory rep to perform the reprogramming, and this too should be made part of international law.

Monday, June 17, 2013

higher fruit


The low fruit in robotics is what we see around us now – most automation is either very limited in scope or else occurs in a controlled context, because even the most sophisticated machines are barely capable of operating in a truly uncontrolled environment, populated with with unknowns. The higher fruit remains largely hypothetical, for the moment.

But, at least as importantly, the low-hanging fruit involves fitting robots into what’s is already happening, as by substituting machines for people doing more or less the same work. Conversely, the higher fruit involves taking advantage of robotic technology to transform what's happening, substituting better practices for those that are currently convenient, or at least conventional. But transformation is, by its nature, disruptive.

I’d like to begin by laying out a conceptual, two-dimensional graph of possibilities, expanding upward and to the right from a point of origin. Let’s call the horizontal axis time, with what’s technically possible now near the left edge and what will sooner or later become technically possible arrayed to the right, according to how far in the future it can be expected to come within our reach. The vertical axis is a little trickier. At time zero (the present), you could call the bottom of the graph ‘current practice’ and points higher along the left edge represent increasing departures from current practice, all possible within the constraints of current technology, but which could be expected to involve greater investment and/or to encounter increasing resistance from vested interests, or other forms of economic or social inertia. Let’s label the vertical axis ‘disruptiveness’.

As you move to the right (applications of technology not yet within the state of the art), data points above the horizontal axis represent applications which, even if they were to become possible tomorrow, would still involve varying degrees of disruption of current arrangements. Current arrangements being financial, legal, social, habitual, or any combination of these, but specifically including current practice as what would be most directly disrupted.

Now, for the sake of simplicity, let’s reduce this graph to a four-celled, two-by-two grid (see graphic), with the lower-left cell representing current practice, the lower-right cell representing predictable advances in technology that, once available, can be applied without disrupting current arrangements, the upper-left cell representing transformations that are technically possible now, but which can’t be applied without disruption, and the upper-right cell representing what will become possible at some point in the future, given predictable advances in technology, but which can't be applied without disruption, unless current arrangements change in the meantime.

That upper-right cell is a kind of Pandora’s Box. It contains many wondrous things that, were they imminent, would nevertheless be characterized as impractical, because they somehow run afoul of current arrangements. Nevertheless, this is where you’ll find much of the potential of robotics.

Take the notion of Personal Rapid Transit (PRT) as an example of something overlapping between the upper-left and upper-right cells. PRT is an application of robotics to transportation which has been knocked about for several decades. Technically, it’s not such a great challenge that a determined effort couldn’t get it working acceptably on a large scale in fairly short order, a few years at most, but that determined effort can’t happen, because deploying such a system would require a large investment in new infrastructure and would work to the disadvantage of those who are already invested in some aspect of automobile transportation, including car owners since that infrastructure would inevitably mean taking some space from existing streets and some funding from their maintenance. Instead we get cruise control and automated lane following in otherwise conventional cars running on conventional roadways, and PRT research lives perpetually on the back burner, making slow progress.

To bring this back around to the application of robotics to agriculture, from a certain point of view it's easy to overlook the disruption involved – displacing monoculture, heavy equipment, and petroleum-based fuel, fertilizer, pesticides, and herbicides – as a net gain in itself. But not everyone would agree, certainly not most of the companies whose primary business is marketing such products to farmers. And most farmers would surely be pushed far outside of their comfort zones if the whole vision of replacing conventional agriculture with what is essentially robotic gardening were to be ready today and imposed by fiat tomorrow. Wherever you find yourself on that spectrum, it must be obvious that the application of robotics to plant care in a very detailed manner, on a plant-by-plant basis, if scaled up to a significant percentage of the land area currently under cultivation, holds the potential to be hugely disruptive. Because of this, I’m not particularly surprised to see research of this sort proceeding on a shoestring, privately, or outside of the US.

On the other hand, its potential for disruption is no measure of the intrinsic value of a potential technology, and some technologies are needed, even desperately needed, despite that they are likely to run afoul of current arrangements. I would argue that the application of robotics to making the best practices of horticulture scalable is one of these, perhaps even the poster-child example.

This is the fourth post in a series: Part 1, Part 2, Part3.

Saturday, May 25, 2013

aligning research for long-term relevance

“Luck is what happens when preparation meets opportunity.”
Seneca the Younger

“Change brings opportunity.”
Nido Qubein

No one wants to spend their life doing something that ceases to matter before they've even reached retirement, yet who can say which enterprises will wither and which will prosper, and for how long. We make our choices and take our chances that fate and fashion will be kind to us.

On the other hand, some things are predictable. One such is the increasing price of petroleum and products derived from it, over a matter of decades, as easily pumped supplies dwindle, as the techniques involved in extracting other supplies become both more expensive and less acceptable, and as calls for a gradually escalating tax on carbon emissions gain traction. Another is that the already high concentration of greenhouse gases in the atmosphere will climb even higher before leveling out, with dramatic consequences for climates worldwide.

Yet another is that given other pressures on agriculture, from climate change and the rising price of fuel and other petroleum derivatives, practices that result in soil erosion and the mineralization of what remains will become even less popular than they are already, perhaps to the point of becoming illegal. In the future, if tillage is to be used at all, contour farming might be required wherever there's enough slope for water to collect in puddles or run off the land rather than percolating into it. But, realistically, routine tillage is a problem no matter how you go about it, ten thousand years of precedent notwithstanding. No-till farming is gaining in popularity – although, at present, frequently depending heavily on herbicides, something else that's becoming increasingly unpopular. There has to be a better way.

So say you're a professor, postdoc, or grad student with an interest in agricultural robotics. To what should you apply your efforts? If you climb onto the bandwagon of the latest fad, whatever it might be, you run the risk that you're work will never be put to practical use. Conversely, if you simply work to automate practices that are currently common, you risk seeing your work devalued as those practices are displaced altogether by some new approach to farming. What to do?

The answer to that will necessarily be largely driven by your technical background and interests, which piece(s) of the complex technological puzzle that is robotics you are comfortable with and/or motivated to work on. Happily, some technologies are broadly applicable, and will be useful no matter what farming system is in use. One such example is UAVs that are large enough to carry an array of sensors and robust enough to deal with vigorous winds; the information they can provide will be valuable no matter what is happening on the ground.

Other technologies are more tied to specific farming practices. This is commonly seen in row cropping, where crops are planted in rows that are far enough apart for the wheels of smaller tractors or self-propelled sprayers to fit between them, with the spraying equipment and implements pulled by those tractors also designed to fit between the rows. But most such equipment is designed for use with narrow rows separated by wider gaps, so just the change from narrow rows (single lines of plants) to wide rows (densely planted swaths, perhaps eight feet wide) renders much of this equipment useless. A more dramatic shift, such as a switch from monoculture to polyculture, especially a polyculture that incorporates perennials, could force us to rethink our entire approach to field work and idle most existing equipment.

That existing equipment represents a huge investment, and therefore its removal from service can be expected to encounter a huge amount of inertial resistance, which is why no change of this magnitude could possibly happen overnight. It will take years, perhaps decades, before new equipment, suited to a new way of doing things, can be designed, factory production ramped up, and existing inventory replaced. Considering that this new way of doing things is likely to parallel or precipitate other changes, for example in the nature of commodity markets, this gradual transformation may be a good thing, as it will provide time to make other adjustments.

What will this new way of doing things look like? First it must be far less destructive of the environment than current methods. There are some objective measures for this, for example: how much soil is blown or washed from the land each year, contributing to the dust load in the atmosphere and the sediment load in streams and estuaries; how much of what remains on the land is organic material (decaying plants and humus); the amount of fertilizer, pesticides, and herbicides that find their way into the atmosphere and into aquifers and rivers; and the loss of biodiversity (plants and animals) caused by the destruction of natural ecosystems to make way for agriculture. Ideally it should be regenerative, gradually healing the destruction already done.

But, at the same time, it must be productive on a scale at least approaching that of conventional modern agriculture, as measured in terms of food value (calories, protein, dietary fiber, vitamins, minerals, etc.) per acre per year. To sacrifice this would be to invite destruction of the last remaining wildernesses to make way for expansion, and/or famine. Some slack in the need for production can be created by a shift away from grain-fed meat and dairy products.

This combination of requirements, that agriculture continue producing food of acceptable quality in adequate amounts, while at least becoming environmentally benign, and without further expansion onto land not already committed to agricultural production, is the minimum acceptable scenario. It is also a formidable challenge, requiring intensive management to make the best possible, gentle use of available land. If that intensive management had to be provided by people, we might need to move upwards of 25% of the population back into agricultural production. Happily, robotics offers an alternative approach, one that doesn't involve forcing people back onto the land.

But, even using intensive methods, how is it possible to accomplish maintaining or increasing productivity without sacrificing environmental concerns? Many believe the answer to this is agroecology, the use of agricultural methods that mimic (and, on the edges, blend into) natural ecosystems, which at the minimum means both polyculture (intermingling of companion plants) and permaculture (the inclusion of perennials). That agroecology should also be free from reliance on routine tillage is not yet so widely appreciated, however one has only to glance at the floor of any mature natural ecosystem to see what nature has to say on the subject. Tillage is best accomplished by worms and other burrowing critters.

Routine tillage excepted, most other gardening techniques can be used, and most (perhaps all) of these can be performed by robotic machinery. Because the qualities that differentiate robotic systems from ordinary mechanical equipment are exactly what the new, intensive approach will call for, and because the scale of agriculture dwarfs most other enterprises, the necessary, if gradual replacement of existing equipment described above represents a tremendous opportunity for roboticists and robotics companies, and a guarantee that work invested to bring it about will remain relevant for a long time to come.

This post is the third in a series. (Part 1, Part 2)

Friday, May 17, 2013

design options, and factors which impinge upon them

This is the second post in a series, which began with what operations should a ‘cultibot’ be able to perform?.

If you've read the masthead for this blog, you'll know that it is about how robotics can assist with the application of horticultural methods on an agricultural scale, so, even though I've been talking about gardening, the real goal is to apply gardening methods (or at least the attentive attitude of a careful gardener) over the entire landscape, ‘from sea to shining sea’, wherever land is managed as opposed to being left entirely wild. That's a tall order, of course, and no single system could conceivably cope with the vast range of conditions implied.

So, to simplify matters somewhat, I'll confine myself to consideration of land primarily engaged in production for market, whether for food, fiber, fuel, or any of the other reasons we put land to productive use, and to systems engaged in the management of land for production.

In this context, it's reasonable to divide plants into two basic categories, crops and weeds. Any such system must be able to differentiate between the two and nurture crop plants while eliminating or at least suppressing the growth of weeds. Differentiating plants is more easily accomplished after they have grown beyond the seedling stage, but controlling weeds that grow from seed is more easily done while they are still seedlings. So one design choice is whether to attempt to differentiate seedlings and eliminate weed seedlings before they've had a chance to root deeply.

Differentiating seedlings can be made considerably easier if you know very precisely where the crop seeds have been planted, so anything appearing more than a very short distance from one of these locations has a high probability of being a weed. Beyond that, a robotic gardener could use visual information (from gross morphology to the vein pattern of the seed leaves) and/or sniffing for volatile compounds. Visual information can be acquired from a considerable distance, relative to the size of the seedling, whereas chemical sniffing requires close approach, and is generally better accomplished by a small device.

Either way, once the decision to take out a seedling has been made, there remains a choice of methods. If the device which detected it is small, light, and locally supported, meaning that its weight rests on nearby soil, then a mechanical extraction of the seedling from the soil (uprooting it) may be the most reasonable approach, the main concern being that that it grasps the correct seedling.

On the other hand, if the device is remotely supported, meaning that it is essentially suspended amid the foliage, mechanical uprooting may still be an option, if it also possesses one or more quick-moving appendages, but the time involved will be a more pressing consideration, as a remotely supported device would necessarily represent a larger investment, so there will be fewer of them and each must handle a larger area. But other options are available to a remotely supported device. It might use a laser which can be retargeted as fast as a mirror can be repositioned, or a miniature turret that fires ice pellets, and if it uses an arm, that arm might slice through the seedling with a thin jet of high pressure water rather than grasping it.

So much for weed seedlings, but they're the easy example. Let's consider weeds, like some grasses and shrubs, that spread by rhizome. A rhizome is a horizontal root, running just below the surface, occasionally sending up stems which appear from above to be independent plants, but which are actually part of a larger, robust organism. If a plot is heavily infested with plants that propagate by rhizome, they cannot be eliminated without also destroying whatever crop was in the ground at the time. The approach that works most reliably is to frequently survey for emerging stems of these plants, pulling them up with as much of the rhizome as comes along fairly easily, and to keep at it until the rhizomes run out of stored energy for putting up new stems and rot in the ground. This is an operation that really can't reasonably be performed by a locally supported device, because it can require significant force to pull up such a stem, much less the attached rhizome. A locally supported device might clip them off at the soil surface, but to be effective, that would need to be repeated even more frequently than the uprooting.

Something a small, locally supported device might do would be to apply very small doses of some specific herbicide, such that it would be absorbed into the rhizome itself before killing the stem, but this begs the question of whether herbicides are to be used at all. (If they are to be used, then direct application of small doses to the weeds to be controlled is the obviously best approach.) Something a remotely supported device could do that a locally supported device really couldn't, would be to use an electromagnetic sensor to trace a signal transmitted from one section of rhizome to locate another spot where it could be exposed and then heated by passing an electrical current through it. Another option would be to inject steam next to the stem, killing the stem and a short segment of rhizome.

If we were to go on to consider other operations, for example harvest and dealing with the coarse plant material left over after harvest, it would quickly become apparent that the larger, remotely supported machines are necessary, whereas the smaller, locally supported devices are probably optional, although the most efficient approach might well be to use both in a complementary fashion, the gardening system mentioned earlier, perhaps even including insect-sized devices that fly among or crawl over the plants themselves.

Strictly speaking, devices suspended from a wheeled platform that always follows the same tracks, a practice which is beginning to gain traction in precision agriculture circles, would qualify as remotely supported for the purpose of this discussion, even though part of the surface of the field is sacrificed to supporting the weight of the machines. Where the wheels of those machines don't run, the soil remains uncompressed. Placing gravel on the tracks they do use would help limit damage. Another option would be to install rails or troughs that double as channels for water transport. Still another option would be to support the platform on legs that only ever placed their weight on particular spots. Any of these options would help avoid the rutting of continuous paths that could lead to erosion.

Other design choices include sensors and the sorts of information gathered and the types of manipulators provided. A garden is a very complex environment, most of which is delicate, and while that delicacy makes the design and control of manipulators more difficult than it would otherwise be, getting adequate information into the machine, and deciding what operations to perform, is by far the greater challenge.

Laser scanners of various sorts could prove very useful in mapping foliage. Imaging radar could help locate stems that are hidden by foliage. Video, particularly if high definition not only in pixel density but in color depth (and ideally in splitting the spectrum more finely than the usual RGB), combined with telescopic optics, could make differentiation of seedlings far easier. Chemical sensing, mentioned above, could be used, for instance, to determine the ripeness of fruit and to detect the presence of some infestations and diseases. Very acute hearing could help with the discovery of insect activity. Infrared video would help with locating warm-blooded animals. Touch would be important in making it possible for manipulators to reach through foliage or pick fruit without damage.

All that said, the most fundamental choice isn't any of the above, but rather what sort of management regime to support. I'll address this issue in the next installment.

Sunday, May 12, 2013

what operations should a ‘cultibot’ be able to perform?

The list of necessary operations might seem pretty straightforward. A robotic gardener or gardening system (more than one type of machine, each specializing in a subset of the overall range of operations) would need to be able to accomplish the same basic tasks as a human gardener: seedbed preparation, even if it's only a narrow hole through mulch and/or ground cover; planting, watering, weeding and/or suppressing weed growth; applying nutrients in one form or another; intervening to control animal pests and diseases if they threaten to become a problem; tracking the progressing ripeness of the crop and combining that information with the weather forecast to decide when to begin harvesting; and, of course, the harvest itself. Also, to get the best possible productivity from the land and the greatest benefit from the investment in a robotic gardener or gardening system, that machine or system should be able to start seedlings in a greenhouse and then move them out to the open field, inserting them through mulch and/or ground cover.

Then, too, there are other operations which are either optional or less commonly understood to be necessary: either collecting leftover coarse plant matter (corn stalks, tomato vines, and so forth) for centralized processing (grinding, optionally followed by anaerobic digestion producing methane gas, followed by aerobic composting) or simply grinding it into small bits in the field and leaving it where it falls, for mulch; limited relocation of soil to create and maintain a topology which slows the runoff of excess precipitation, without creating pools that persist hours or days after the latest storm has passed; relocating larger stones found in the soil to create permeable dams across grassed drainage ditches, where they can help slow water movement and help prevent it from cutting through the sod; and dredging muck from the bottom of surge ponds and distributing it over the field, to recover nutrients and prevent them from overloading any permanent ponds, much less passing downstream.

But, if you're going to have robotic machinery doing all of this, there are other operations you might want it to perform, operations which machines that have sensors and processors and well as effectors can relatively easily be made capable of performing. Examples might include: keeping detailed records including when each seed was planted and what the soil temperature and moisture content was at the time, when the sprout emerged from the ground, and how fast it grew; monitoring each plant for early signs of infestation, disease, nutrient deficiency or toxicity, waterlogged soil, and so forth; and targeted pruning to control infestation and disease without drastically reducing leaf area.

In the next installment, I'll begin to address design options and mechanical details relating to making machines capable of performing the basic operations listed above.

Monday, April 15, 2013

In agriculture robots replace job vacancies

Here's the bottom line…

So as new types of machines find their way into the fields, rest assured that they are not, for the most part, displacing workers who would otherwise be in those fields, but rather, in some cases, moving them into more technical work as robot tenders, and in other cases taking over work that fewer and fewer people are willing to do for the money it pays, and that, for those few who are displaced, there will be other farmers nearby anxious to hire them. Meanwhile, a new industry will be germinating.

See the full article on

Thursday, March 28, 2013

What's in a name: sustainable agriculture vs. agroecology

Names are important in much the same way that book covers are important; they suggest what lies behind them. But there is an additional way in which names are important, which is their resilience in the face of attempted cooption. Does the chosen name continue to robustly represent what it was originally intended to represent, or does its meaning become diluted by misapplication?

There are quite a few names available for use in referencing the practice of growing food (including herbs and spices), flowers, animal feed, forage, fiber, fuel, lubricants, and chemical feed stocks, all in harmony with nature, most of which take the form of an adjective followed by one of the following nouns: gardening, farming, horticulture, agriculture, viticulture, etc. The set of adjectives in use in this manner includes biodynamic, biodiverse, organic, regenerative, resilient, and sustainable. (To complicate matters further there are also some terms relating to specific practices, like permaculture and polyculture, that are important to any such discussion.)

For many of us, there is a tendency to use adjective-noun combinations from this set almost interchangeably. We understand what we're attempting to invoke better than we understand the nuanced distinctions between them, so it seems to make little difference, except that this results in a profusion of terms for essentially the same thing, and some of these term combinations are more susceptible than others to being used in ways that lack clarity or are even contradictory to what we would mean by them, meaning that there is a danger that others hear us saying something different from what we meant.

“Sustainable agriculture” is, unfortunately, one such term. While it originally meant something like ‘a set of practices which can be continued forever without wearing out the soil or reducing yield’ it has come to mean other things to other people, including something on the order of ‘a set of minor adjustments that allows agribusiness to continue doing essentially what they're doing now for another decade or two without catastrophic collapse’.

Another term which is more difficult to fully comprehend, but less susceptible to misuse is “agroecology” (“agroecological” as an adjective). Properly understood, it implies all of the other adjectives listed above, except some of the more arcane aspects of biodynamics, and adds one important concept, cooperation between human-managed production and the native ecology of the land in use, such that you can't tell where one ends and the other takes over.

Just as there is a confusion of terms with respect to the practice of managing productive land in harmony with nature, so too there are many terms for robots designed to assist in this process, nearly as many terms as there are robots of this class, it would seem. Rather than slog through the list, I'll get straight to the point, which is that, after years of using “cultibot” and “cultibotics”, I've found another pair of expressions that I prefer. “Agroecological robotics” nicely defines the field in a manner that resists dilution, and “agroecobot” works well enough in reference to actual machines. I'll probably continue to also use “cultibot” and “cultibotics”, but understand that I intend them only as shorthand for machines designed to assist in the practice of “agroecology” (which itself implies biodiverse, organic, regenerative, resilient, and sustainable practices), with a view to making that practice scalable to millions of acres.

Saturday, February 02, 2013

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.