Farmbots Have Mechanical Green Thumbs

Is there any robot better than roomba.   You press go, it cleans the carpet!  Well, as it turns out, the unfortunately named: SUFFER’11 farming robot is better.  You press go, and it plows the field.

Developed by the University of Southern Denmark, the SUFFER’11 is a robot designed to tackle some of the common repetitive tasks common to farming.  Physically, this bot is nothing more than two treads, a CPU, and a center attachment for swapping farming implements.  Anything that attaches to the back of a tractor, can hook to this bot (plow, sprayer, harvester)  The SUFFER’11 is both remote controlled and has a camera module that is able to detect rows, and navigate accordingly.  

For an industry that is becoming increasingly mechanized, with GPS guided tractors , and automated irrigation systems, the autonomous farming robot seems like the appropriate logical progression.  

Dear Google: please add this feature!!!

 Dear Google:

I love android because it is so customizable!  That aspect really plays to the geek in me.  Also I hate android because it is so customizable!  As it turns out, I lack the artistic ability to create simple, attractive, organized layouts.  The perfect target for your devices is the tech savvy graphical designer (not a large demographic.)  Those who are tech savvy, but not artistically inclined, or artistic without a solid tech understanding are left to limp through your interface, created underutilized or simply confusing GUIs for ourselves.

Downloadable profiles would solve this.  I want the ability to click on a profile that was designed by someone else and have it configure to my phone.  This profile would need to be able to download its own apps (with my permission of course).   

For example, I would love to have a profile of Google staff picks, or a Gizmodo profile.  As a golf lover, I would download the golf digest profile, and their vetted and approved golf gps / golf training / golf gear / and weather apps would be automatically downloaded, and organized on my phone.  

   

Android is too complicated for those who lack tech sense.  There is no way that I can get my mom an android device.  It’s too complicated for her.  But if I could download a dog lovers / worlds best grandma profile she would never need be confused over how to setup her phone, or what apps she should download.  

This is the perfect solution to partner skins.  If HTC sense, or moto blur wants to load a skin, that’s fine, you can load the HCT sense, or moto blur profile.  That way when the next version of android comes out, the users would not be at the mercy of HTC and Motorola for updates.  We could updates our OS, and their profile would either work or not.

Enterprise customers could load a corporate profile with the apps specific to that company’s databases and systems.  The security problems for enterprise customers with mobile compute platforms, is not a new issue.  Its one with two parts: apps that are secure, and phones that are setup correctly.  The app building process is being tackled by every large company in America right now.  But the ability to roll out these apps to their employees in a secure fashion is a problem that currently requires personal IT support for every device.  Downloadable profiles could fix this.  Your IT department only needs to setup the profile once, and the profile rolls itself out to the employee base.

Downloadable profiles would enable the rest of the world to enjoy ergonomic, well organized layouts, even without the artistic or technical talents, that great UIs require.

Challenge: build a home using robots.

Robots are great at simple repetitive tasks, which work out great for an assembly line, but bad for construction.  The problem with home building is that there are many separate tasks, some of which require specific onetime actions. The different tasks for homebuilding include:

• excavating
• Building a level concrete foundation.
• Framing
• Electrical wiring
• Plumbing
• Drywall hanging / finishing
• Roofing
• Bricklaying / Stone masonry / vinyl siding.

Lets take a single simple job and see if we can break it down to a robot task:  Framing.

This job starts with cutting wood.  For simplicity sake lets say that stacks of wood have been delievered and organized such that the robot can find the right piece of wood (in this case an 8’ 2x4), maneuver it to a cutting area, measure the board, then make a straight cut at the desired mark.  This job could take years of programming a humanoid, or could simply take an automated forklift, and a CNC table.  The folks at wikihouse are working on implementing this very thing:

Now we have the correct size board in hand, we need to maneuver it to the correct position on the house.  Yet again, a simple job for a human, but one that requires several steps for a robot.  First, move the wood to the correct place on the house.  There are several ways to do this including a robotic arm on your automated forklift, but this has issues with upstairs placements and nonstandard positions.   Instead, lets say that you have a team of quadrocopters for this task.  Check out the teamwork achieved by several quadrocopters to build a tower in this video:

 

Next you need a way to stabilize the board and hit it with a mobile nail gun.  This is tricky, because you can’t think about the project as a simple wall, rather it could be a complex architectural shape with non square angles, and difficult to access joints.  Let’s face it, every home built in the past 50 years has these features.   For this task lets use two motorized clamps connected by a motorized telescoping rod.  Connect the clamps to the board and its new surroundings with the quadrocopter, then let the telescoping rod expand or contract until the board is in the correct position.  This board can then be tacked into place by a nail gun mounted on a separate quadrocopter.

OK lets say you are a critic who thinks that quadrocopters can’t lift heavy things like boards and cant fly under things.  Specific to flying under things, you are correct.  A different mode of building would need to be used where houses are built from the ground up.  As for the weight limit check out this beauty, and think teamwork!

The original question was can a robot build a house, and the simple answer is no, or more correctly, not yet.  However, can a team of specifically designed robots build a house?   Perhaps!

IBM declares post PC era has begun!

IBM CTO, Mark Dean stated in a Blog post last week that “PCs are going the way of the vacuum tube, typewriter, vinyl records and CRT.”  That the age of the PC is over and the post PC era has begun.  These are big words from a credible source.  Since transitioning out of the PC hardware business, IBM has surpassed tech giant Microsoft in market capitalization.  So, is he correct?

Certainly the past three years have been a boon for mobile computing.  The tech industry has seen smart phones with dual 1Ghz processors, tablets sales dramatically cutting into laptop sales, cloud computing is on the rise, and network based browsers are growing faster than local browsers.   Thanks, to leading edge process nodes we can make more powerful electronics in smaller packages than ever before.

But in all this technology invention, there is one thing we have not been able to change or make smaller: the human users.  Netbooks taught us a lot about computing ergonomics.  First, there is an optimum keyboard size, that is intrinsically tied to the average size of a human hand.  Second, Netbooks also taught us about screen size and resolution: to usably view the internet you need more pixels than WVGA screens can provide.

What about the younger generation?  Sure, a person in their 30s might not find iphone screens as usable as PCs, but that is because they grew up with PC.  Kids, who grew up with iphones should have no problem making the paradigm shift, right?  Not quite.  A study done in 2009 compared typing on a blackberry, a laptop, and a full size keyboard.  The study found that people could type on the blackberry for 5 minutes before become fatigued and irritated.  On the laptop that number went to 30 minutes.  And on the full sized keyboard the time to fatigue was 3.5 hours.  Although not included in this study, I would speculate that hunt and peck touch screen input would fall short of the 5 minute blackberry mark.

What makes a compute platform a PC?  A full size monitor, a full size keyboard, and some type of CPU.  Although it is true that many of the processor and memory intensive activities of a PC are being moved to servers rather than rely on your local CPU, the things that make up a PC are not going away.  

source

If China is the next America, who is the next China?

In the 80s the US began wide scale outsourcing to Malaysia, Thailand, and the Philippians.  This outsourcing proved extremely profitable for both the US and our Asian partners, until the tax subsidies expired and the labor rates naturally increased.  At such time, the US fell quickly out of love with south Asia and in love with China; a land of plentiful low cost labor, and thanks to the free trade zone in Gongdong, virtually no taxes.

In the past few years we have witnessed increased labor costs in china and experts have begun to look for a new land of low cost manufacturing.  In the short term China is doing well to protect itself from these labor prices by moving manufacturing to more remote locations, where the cost of living is still low; but this is just a stop gap measure.  As the cost of living increases, so must the wages.  Some experts believe that Vietnam is the next China.  Others believe that Brazil is an ideal candidate. 

Personally I think that China is the next China.  Wide scale move from china is not going to be as easy as moving from the Malaysia was 10 years ago.  The reason is the supply chain.  If you move the final assembly of a product from China to Brazil for example, you will still need to manufacture most of the subassembly components in china and ship them to Brazil for assembly.  Lets not forget that China controls 95% of the rare earth metals.  As an example, lets imagine that Apple wants to move the iPhone manufacturing from Foxconn in china to Brazil.  The micro processors are made in Taiwan on a world leading 28nm fab.  This fab can’t be moved to Brazil without 10s of billions of dollars and 10 years investment.  So for now they must be shipped.  The LCDs are made in either Korea, Taiwan, or China with rare earth materials that come from only china.  The li-ion batteries also have to be made in china, because their materials are only found there.  Now rather than paying the shipping for one boat to bring a completed phone from Asia to the US markets, you need nearly a dozen boats each bringing different components to South America, then One more ship to take them to the US distribution chains.  And you need to increase your inventory management capability in Brazil, which results in longer lead times, more expensive components, and waste.

If manufacturing can't easily be moved then what will happen?  Will manufacturing just stay in China and everything be more expensive?  Terry Gou founder and chairman of Foxconn has a solution: robots.  Last week he announced that the company would begin replacing his human work force with robots.  He has plans for adding 1 million robots in the next 3 years.  When you realize that is current manufacturing workforce is comprised of 1.2 million workers, this represents almost a complete overhaul of his empire.

After a year plagued by worker suicides and falling profit margins, Foxconn has taken strong steps to controlling the labor variable in the manufacturing equation.  One has to imagine that this will not sit well with the ruling labor party of China, but the truth is, as long as first tier manufacturers can maintain their profitability, the entire support network (not run by robots) will also thrive.  Thanks to the outrageous success of first tier manufactures like Foxconn, Shenzhen China is fastest growing city in the world, home to 25 million people.  Although the first thought of robots replacing human workers may seem unpalatable, there are 25 million people in the city of Shenzhen whose only economic hope is the success of these bots to keep the city and their region in the black.

See source link :

http://www.appleinsider.com/articles/11/07/31/foxconn_to_substitute_workers_with_1_million_robots_in_3_years.html

Robots Should do EVERYTHING

I have one over arching goal when it comes to technology: Robots should do everything.  The next time you are mowing your grass think about that.

This belief is not based in a desire to never mow the grass again, in fact, I enjoy mowing the grass.  Rather it is motivated by a desire to be free.  You can mow the grass if you want to, but you don’t have to if you don’t have time.  Or at a larger scale, lets say your yard is 20 acres in size, and you are 70 years old.  All of a sudden, a robotic lawn mower becomes not a tool to get out of work, but a key requirement of your housing freedom.

On a less robotic scale, this mentality applies to all aspects of life.  If you need creativity to solve a problem once, develop a process to solve this same problem when it happens again.  This new process frees up the creative to do other things, and solve other problems, without becoming bogged down in repetitive execution.  For example, if you work at the post office and half the people coming in just want to buy stamps, then an automated stamp vending machine, has just freed you to be more productive and useful, without the repetitive dolling out of stamps.

Every time I see manual labor, I can't help but think: how can a robot do that same job?  Take for example home construction projects.  A CAD file is created by the architect, this is printed in 2D for the general contractor, it is interpreted and explained to specialists who cut boards, frame walls, hang drywall, finish drywall, paint.  Its difficult if not impossible to create a robot with the versatility of a human, but it is completely possible to supplement the specialty laborers with specialty robots.  You might not be able to create a roofing robot, but how about a robot that can place shingles,  and a different robot that can nail them in place.  One robot might not be able to build a house, but a group of robots could build a subdivision.