Robots of the future will have even more capabilities than the ones are currently being tested.  Hopefully, the final outcome of endeavors in rural robotics is that human involvement in the entire process of planting, growing, and harvesting crops is nearly supplanted.

Currently research teams are developing various robots, including those that:

• Harvest Crops [13]
• Remove Weeds [6][10]
• Water Plants [8]

 In the future robots will even be able to handle other factors of plant growth such as:

• Regulate heat and temperature
• Prevent plant disease
• Maintain sufficient light exposure
   

Essentially, the above mentioned six capabilities are the major features of a greenhouse, a facility that isolates and controls a field of crops to grow.  With regards to temperature control, current greenhouses have covered glass sun roofs that absorb and trap heat from the sun [3][4].  To apply the same concept to wide crop fields, similar insulating material will have to be developed so that it is cheaper to build sun roofs that cover larger areas.  The result is an extremely large green house, one that has high labor costs if humans were assigned to maintain it. That’s where robots come in.

Features These Robots Will Need

To begin with, these greenhouse machines will automatically detect whether there is too much or too little heat suitable for growing through the use of thermometers and other devices.  They will then produce or reduce the right amount of heat to compensate, using natural sources of energy such as chemical reactions or solar power. 

Similar concepts can be applied to regulating humidity and moisture, where robots will release calculated amounts of water vapor into the giant greenhouse.  Because moisture and heat are now controlled, robots only have to use a set amount of water for the plant, independent of precipitation such as rain or snow.  Humans will only have to check the robots several times a year. Ultimately, crops will no long be grown according to season or geography, as the major factors of heat, moisture, and precipitation will all be controlled. 

Robots within these greenhouses will also harvest the crops. Production will certainly increase because the robots will operate night and day.  Future robots will also prevent plant diseases.  Normally, people assigned to apply pesticides have to wear equipment is not only uncomfortable for the applicant, but also puts the worker at risk of overheating and dehydration [2].  Replacing the human with robots will potentially eliminate health risk of greenhouse workers.  Robots will also make use of more eco-friendly measures to prevent plant atrophy, without the use of chemical pesticides.  Already, there are rural robots that autonomously and manually remove weeds, a common nuisance to plant growers.

Technology Advances Needed

• Sun Roof Material
  As mentioned above, current greenhouses have glass sun roofs to partially control the temperature [3].  However, if the same idea were to be applied to a large scale, say a crop field, glass would probably be too expensive.  Lighter, cheaper materials like semi opaque plastic would have to be developed and used as the key material to build the expansive sun roofs [1].
   
• Eco-friendly plant disease prevention methods.
  Currently, the major plant plant being dealt with by robots is weeds.  Both Lukas and the Illinois University model manually remove weeds instead of using harmful chemicals [6][10].  Plants should be protected by other toxins and nuisances, not just weeds.  Although many farmers simply use pesticides and other chemicals to prevent other diseases, these are neither safe for the applicant nor the environment [2].  In the future, eco-friendly methods will be invented to immunize the plants from other diseases.
   
• Image Perception
  Unlike humans, robots not only have difficulty viewing real objects, but also have a hard time perceiving what they see.  While some robots like the Illinois University weed robot claim to have "20/20" vision, other robots including Lukas can differentiate crops from other plants only in the best lighting and weather conditions [6][10]. Without a better sense of sight and observation, robots in agriculture are in danger of damaging plants, misinterpreting plant toxins as the actual crops, or even harming other robots.

Our Ideas

Among the factors that determine plant growth, light is the most important.  With the exception of fungus, all food grown on farms depends on the sun.  Evidently, too much or too little sunlight is detrimental to any crop. 

With regards to sunlight deficiency, there is little we can do with robots to create light because, while heat and disease prevention can be artificially generated, most often times the best source of light for plants is naturally generated from our sun.

In the discussion of overabundance of sunlight, robots can potentially offer a solution by providing shade.  Here, robots will analyze data given by light sensors and then shade the robot by a certain amount.

To make this project interesting, we decided to create a working model of a automatic shade that would control the amount of sun light reaching a plant.  We are building this model out of a Botball kit from 2006.

The Gameboy/XBC has light sensors for input and servos to control the shade. 

Here is our implementation. First the shade is closed (too much sun), then the shade is open (cloudy).

We are certain we can immediately control the position of a shade based on the current brightness of the sun.  But we believe the plant grows best when the cumulative amount of the sun is controlled.  For example, if it is sunny all day, we may want the shade partially closed in the afternoon.  But if it is cloudy in the morning, we may want the shade open all the way in the afternoon. This is a more difficult program to create.

The shade will be adjusted on the current light and the recent history of lighting to maintain a constant amount of light over every day.  The program below measures a summation of light through a loop.  Once the the total light absorbed reaches a target amount, the servo will close. Then the robot will reset the total amount absorbed back to its initial value and start a new cycle. Future developers may take the program a step further: once the shade is closed, the plant light sensor will then measure when the shade has been closed for too long. 

Here is our code.

Although the program is not fully developed, the general principle is same for the XBC and actual agrobots.  In real life, the shade itself will not be completely opaque.  And because plants can only absorb precise wavelength of light, the shade will have to change its opacity, like a transition, so that the crop receive the right type and amount of light.