ABB – A Waking Tiger

Last week, at RoboUniverse, one of the largest suppliers of industrial robots was noticeably absent, ABB. This powerhouse of automation that has installed over 250,000 robots worldwide could be resting on its laurels, but the latest news from Zurich suggest otherwise. Earlier this year, ABB declared war on Baxter and Sawyer with their new “YuMi” collaborative robot (see video).  Now this week, ABB announces its new robotic venture accelerator in the UK and the opening of its first US manufacturing plant.

 

ABB’s investment accelerator, called IdeaHub, is currently offering students, academics, entrepreneurs and businesses the chance to collaborate with ABB’s team to take their robotic innovation to the next level (and feed ABB’s R&D pipeline). IdeaHub is currently accepting applications, and will start its first class in July. To apply, the accelerator is asking each applicant to address three challenges facing the industrial robot market:

1) Simplicity –  “How will the next generation of robots interact with humans?” – By 2017 the number of installed industrial robots worldwide is expected to exceed two million and they will be working increasingly closer with people in their workplaces and homes. This presents a huge opportunity for innovation as robots come to new industries, to new environments and work with a wide range of people – both trained and untrained.  YuMi is just one example of how ABB confronts this issue.

2) Deep Learning – “How will the industrial robot of the future learn from its experience, or from that of its robot colleagues, to make decisions and do things differently?” – This second challenge seeks to find a way to incorporate more artificial intelligence into the future designs. In the words of ABB, “robots are moving from making cars to driving them.”

3) Ecosystem – “How can we move to the next level, where productivity is multiplied?” – This final challenge reflects the fact that in today’s connected work environments, robots operate within a commercial and technical ecosystem of infrastructure, sensors, accessories, technicians, operating systems and software. This requires an ecosystem where industrial robots that traditionally work in line, step out and collaborate with each other and their human colleagues in new and different ways.

Reading the above descriptions it is almost like hearing Rodney Brooks speak about Rethink Robotics. Curiously, ABB’s accelerator is not being run by their corporate venture capital arm, but by Venturebright, a growth consultancy.

In the words of Venturebright’s operating manager, Tanvir Mufti, “We specialize in bringing together entrepreneurs, start-ups and corporates. While we think the VC funding model is great, there are other things that entrepreneurs might be looking for, aside from just funding. Some may be looking for access to technology to build their product, for customers willing to test their products or for access to markets to sell their product. These are things a corporate can often provide and we use the IdeaHub platform to make connections and then translate between entrepreneurs and corporates in order to make really interesting things happen for both.”

yumi-ABBpx

Ulrich Spiesshofer, made a big splash on Wednesday when he cut the ribbon on its first plant in the United States to manufacture robots for automakers and other industries. This latest move is definitely a judicious decision, since the US constitutes the company’s largest market with $7.5 billion of sales. ABB has been trying to enhance its footprint in the American soil since 2010, and has spent approximately $10 billion in local R&D, capital expenditures, and acquisitions.

According to Spiesshofer, in an interview with The Detroit News, “the U.S. is still the largest economy in the world. It is prospering. It has a great phase ahead of it. The investment in robotics in the U.S. by ABB is a very conscious investment.” Spiesshofer further says Detroit’s automakers have “regained traction” and “found a way to compete,” confirming that General Motors, FCA, and Ford Motor Co., especially, are customers. They “are leading again in certain aspects in their home market, and we want to help them.”

abb car

ABB represents a new kind of manufacturing business on American soil — robotics — that Spiesshofer, a German national, insists is a way to create jobs, not kill them. In fact, this Auburn Hills plant will employ 1,000 new workers.

“Germany would be dead today if it hadn’t invested heavily and continuously in automation,” he says. “Fact is, the countries with the highest robot density in the world have the lowest unemployment rates. We don’t offer robots to take away jobs. We offer robots to … enhance competitiveness.”

The operative word in Spiesshofer’s statement is competitiveness, it’s time to take notice, and not just the startups like Rethink and Universal Robotics.

RobotUniverse lands ln New York, and the World

This week, on the way to the first robotic conference in New York City, I felt a wave of déjà vu bringing me back to the early days of the Internet when trade shows were held in the dungeons of the Javits Center. Riding the escalator down, the excitement was palpable among the hundreds of leading innovators, scientists, and investors gathered. For my readers that missed this historic event, below is a quick recap:

The fiprinted dronerst evening, we held a startup competition that included a wide gamut of the burgeoning industry, from soft robotic grippers to autonomous construction drills to sensor programming platforms.  The big winner of the competition was Voxel8, a novel prototyping machine that prints a matrix of materials such as “thermoplastics and highly conductive silver inks enabling customized electronic devices like quadcopters (above), electromagnets and fully functional 3D electromechanical assemblies.”

The next day Colin Angle, CEO of iRobot, addressed the audience. iRobot is one of the most successful public robotic companies to date, with its fleet of robotic vacuum cleaners, mobile telepresence devices, and military PackBots. Since launching Roomba in 2002, over 10 million units have been sold worldwide, creating a $3 billion robo-vacum market. Now that its robots have been adopted by millions of households, Angle believes they are ready to map out our lives (or homes) while simultaneously sucking up dog hair. According to his presentation, these maps will be able to infer our “predictable behavior” to connect to other IoT devices and anticipate our needs. Yes, one day soon, you will come home from a hard day’s work to have a cold one delivered into your hands à la Roomba.

Following Angle, his former CTO, Rodney Brooks of Rethink walked on stage. Introduced by MC Francis Rabuck as the godfather of robotics, Brooks titled his presentation as “Robots Among Us” (while debunking Hollywood’s techno-paranoia). Clearly, Baxter and Sawyer, are working “among us,” leaving their industrial cousins in their dark cages. These machines are part of a new line of cooperative robotic arms with grippers to speed up production, warehousing, and distribution. Brooks reminded the industry that it is important to deliver on the promise of efficiency to enter the next phase of revolutionizing bots for small and midsize businesses with automation. After his talk, one could walk the small (but well curated) exhibit hall to meet Baxter firsthand, along with Universal Robots’ product line. On the last day of the show, it was announced that Rethink’s competitor, Universal Robotics, was acquired by Teradyne for $350 million (a nice endorsement to the nascent market).

google.carx299_1Following the morning’s keynotes, I walked over to hear a workshop on Driverless Cars. While there were more chairs than people in the room, the quality of my fellow participants was impressive (I even met an inventor of a flying car – chitty, chitty – bang! bang!).  While I have written extensively about the companies pursuing the next wave of automobiles, these panels were able to break-down the variety of platforms being deployed by Google, Mercedes and others. On the left is a screen shot of the Google Chauffeur brain, making realtime decisions using LIDAR sensors, radar, and visioning systems.  In addition to the input into the ECU, we discussed Vechicle-2-Vechicle and Vechicle-2-Infrastructure communications.  There were a lot of big questions that were left unanswered, such as security and regulatory approvals. I hope that by the next RoboUniverse, we will hear directly from Google’s Chauffeur team in addition to other car companies. While all of us were taking bets on the date of mass deployment, Shahin Farshchi of Lux Capital reminded us we are already driving semi-autonomous vehicles today. The critical need for automation in the transportation space was underlined later that evening by the tragic Amtrak derailment.

The first day was capped off by Stephen Wolfram and his computing system. There are few times in your life when you meet an actual genius on the level of Hawkings and Einstein, this was one of those times. Wolfram is more than an physicist, he is a philosopher that is molding our AI future. His software Wolfram Alpha Pro is a monumental shift forward in computing. However, what I enjoyed most about his keynote were his thoughts on a “post-linguistic” world when machines will be using sensors and AI to describe things back to us. Hawkings and Musk decry the fears of AI, while Wolfram embraces its “randomness” to discover new galaxies (below is a recent TedTalk that was very similar to our discussion).

 

The next day we heard from the new dean of the University of Pennsylvania’s School of Engineering, Vijay Kumar.  Kumar, whom I met several years ago, is an innovator of drone swarm behaviors. Besides the whimsical video below, Kumar’s lab is using swarms for real-time disaster recovery and precision farming.  The concept is very simple, in fact, it is inspired by nature (e.g., birds and ants).  His “follow-the-leader networks” are able to make real-time adjustments to work simultaneously to achieve goals to help farmers prepare for the upcoming harvest and first responders deal with the scope of a collapsed building.

 

Similarly, Robin Murphy of Texas A&M and the Center of Robot-Assisted Search & Rescue spoke later in the day about robot teamwork in aiding first-responders.  Similar to Kumar, she deploys robots in the field from the recent earthquake in Nepal to the massive terror attacks of the World Trade Centers (and all disasters in between). Murphy gave the audience very practical advice in understanding the financial and professional motivations of disaster/recovery administrations. One of her most successful examples was using water drones to open up the ports in Japan after the Tsunami in four hours, instead of the six months with human divers.

 

Learning from nature for our mechanical innovations is not novel, in fact Da Vinci’s sketches often reflect similar ideals.  Most people think of Big Dog, however, the Festo Corporation has been using “bio-mimicry” to find real solutions to human problems for years, from grippers inspired by a chameleon’s tongue to propulsion systems that mimic a kangaroo.  Hearing Nuzha Yakoob of Festo describe the team work of ants, echoed the earlier thoughts of Kumar and Murphy. I left the room feeling hopeful to where our robotic industry is headed.  In the words of the closing speaker, Daniel Theobald of Vecna, it’s time to stop building and start selling. We will always innovate but getting the products out to the market to make a real difference will be the only way for us to cross the chasm to mainstream adoption (and acceptance).

To register for Richard Erb’s next RoboUniverse conference in Seoul this June, go to MecklerMedia’s RoboUniverse’s website. 

Inside the Navy’s Robot Testing Center

There is nothing like seeing something firsthand.  As protestors sign petitions and organize movements outside the UN to Stop Killer Robots, the Navy is focused on something entirely different (at least according to The New York Times).

The NY Times has been running an online two-part feature called, Robitica. Above is the installment about the US Navy’s research facility in San Diego (notice the absence of Terminator).  At this facility civilian engineers work alongside active-duty troops to develop and test the next generation of military robots. The engineers are members of the Unmanned Systems Group at Spawar, or Space and Naval Warfare Systems Command, a research and operations arm of the Navy.  Their mandate is simple: Take the soldier out of the minefield.

When autonomous systems are deployed, engineers at the center say they will revolutionize the way the military fights. They envision a day when one soldier will control an entire fleet of driverless trucks, or a driverless vehicle will make a road safe for a Humvee full of troops. They could also assist in detecting and combatting chemical or biological warfare.

Chris Scrapper is leading a team of engineers who envision an autonomous future. On a recent afternoon, they were tapping away at computers to analyze data from a failed run with RaDer (it stands for reconnaissance and detection expendable rover), the boxy black vehicle they’re trying to make drive on its own.

“It’s hard to say when autonomous technologies will be ready for use in combat,” Mr. Scrapper said, adding, “It depends on the threat level.”

As featured on this blog, remote-controlled unmanned robots have been in use by the military for over a decade. IRobot’s Pakbot defuses bombs, armed drones track and strike their targets and the MK-18, an underwater torpedo-shaped vehicle, mimics a dolphin’s sonar to locate mines on the ocean floor. What these robots have in common is that there is one person directly controlling them.

Mr. Scrapper and his colleagues see the future of combat as using fewer humans to control more machines. While there will always be a human operator involved, they say, that operator may be in touch with several autonomous devices at a time.

Mr. Scrapper says that the technology they have developed is “mission-agnostic and platform-agnostic,” meaning that the same technology that makes a Humvee autonomous could be incorporated into a boat or a bomb-defusing robot.

According to the NY Times, Mr. Scrapper says his engineers are not working on weaponizing autonomous robots, however their technology could be used for that purpose in the future (get out the petition).  For now, his goal and the one funded by the Office of Naval Research, is to make a tool that keeps troops out of harm’s way and frees them up for tasks that require human ingenuity and imagination.

killerrobots

A fine ethical line exists between the mission of Mr. Scrapper and the paranoia of the “Stop Killer Robot” movement.  The big question on everyone’s mind is if we witness another Gen. ‘Buck’ Turgidson moment.

Little Bots Do The Darndest Things

One of my fondest memories of hiking the Costa Rican rainforest are the “leaf cutter ants.”  These minuscule inspects are able lift objects 100x times their weight.  The line of worker ants goes on for tens of yards across the jungle floor.  This week roboticists at Stanford University demonstrated the new leaf cutter ant of robots (see video below).

These miniature robots, known as MicroTugs, can pull up to 2,000 times their own weight, and rely on sticky foot pads that give them immense grip.  While the current models are so small they weigh only 9 g (equivalent to one third of an ounce), 12 g and a tiny 20 mg, development is taking place to enable the creation of larger robots that can carry huge weights. With the same proportionate strength, a robot weighing the same as an average human could pull along a blue whale.

MicroTug4

Industries such as manufacturing and distribution could benefit, as robots pull giant loads around a warehouse and place them wherever needed. Construction sites could have tons of bricks and other items carried by bots. For example, a robot weighing in at only 12g yanks along an astonishing 24kg, some two thousand times its own weight.

There is obvious military potential as armies could rely on munitions being taken wherever needed. Likewise, emergency services could use robots to carry a heavy item such as a ladder into a dangerous situation to help rescue someone.  One of the cleverest things about the robots is that with their sticky feet they can even walk directly up a glass wall.

In a demonstration this week, the tiny 20 mg robot was able to pull along 500 mg, and the slightly larger 9 g (one-third ounce) robot carried over a kilogram (2.2 pounds) vertically up a glass surface. Meanwhile, the 12g robot was able to pull along over 24 kilograms.

David Christensen, one of the Stanford Ph.D. students on the project, said the researchers would next look at how to make a number of the robots work together to carry a heavier items, while also “scaling the technology up to larger bots” so they are able to carry very heavy goods.

The project, which was initially exhibited with fellow Ph.D. student Elliot Hawkes’ thesis at the TED conference in Vancouver earlier this year, relies on “controllable adhesive technology” that mimics some of the techniques used by geckos and ants to grip a surface. Rather like geckos, which use microscopic hairlike features on their toes to connect to a surface, MicroTug robots have miniature rubber spikes that cling to a service yet can be easily detached to lift the foot again.

Once again this proves that geckos are not just to save you 15% or more on your car insurance.

Robot Migrant Workers

As we enter the political season with such notable candidates as Hillary Clinton and Ted Cruz, Americans will once again accuse hard working illegal immigrants as the ire of their economic pains.  As a son of an illegal immigrant, who escaped the Nazis as the soul survivor of his family, I take great offense. The truth is that since 2007 illegal workers are down to such a degree that farms in the Southwest are now employing robots.

 

Replacing migrant workers is Agrobot, a 14-arm automated harvester that easily wheels through rows of strawberry plants on farms in Califirnia.  Harnessing high-powered computing, color sensors and small metal baskets attached to the robotic arms, the machine gently plucks ripe strawberries and ignores the unripe fruit nearby.  Such tasks have long required the trained discernment and backbreaking effort of tens of thousands of relatively low-paid workers. But technological advances are making it possible for robots to handle the job, just as a shrinking supply of available fruit pickers has made the technology more financially attractive.

“It’s no longer a problem of how much does a strawberry harvester cost,” said Juan Bravo, inventor of Agrobot, the picking machine. “Now it’s about how much does it cost to leave a field unpicked, and that’s a lot more expensive.”

The Agrobot costs about $100,000 and Mr. Bravo has a second, larger prototype in development (with 60 arms!). Other devices similarly are starting to assume delicate tasks in different parts of the fresh-produce industry, from planting vegetable seedlings to harvesting lettuce to transplanting roses.

agrobot

Farmers of corn and other commodity crops decades ago replaced most of their workers with giant combines and other machines that can quickly cut and gather grain used for animal feed, food ingredients and ethanol. But growers of produce and plants have largely stuck with human pickers—partly to avoid maladroit machines marring the blemish-free appearance of items that consumers see on store shelves.

An abundant supply of workers, particularly from Mexico, willing to plant, pull weeds and harvest ripe crops for relatively low pay also had suppressed the need for mechanization. But the number of unauthorized immigrants in the U.S. workforce has been declining since its peak in 2007, according to the Pew Research Center, in part because of increased job opportunities in Mexico, as well as tighter U.S. border patrols.

With workers in short supply, “the only way to get more out of the sunshine we have is to elevate the technology,” said Soren Bjorn, Americas unit head for Driscoll Strawberry Associates Inc., the country’s largest berry brand. Driscoll’s largest berry grower, Reiter Affiliated Companies LLC, is partly financing the development of Mr. Bravo’s Agrobot.

Robots have their own drawbacks. They need maintenance and repair—Agrobot normally has 16 arms, but when the Wall Street Journal visited the farm two were recently on the fritz. Some farmworker advocates worry that increased mechanization could also help eliminate jobs that are still needed. And others fear it will give an additional advantage to large growers that can afford to invest in the latest equipment.

Proponents say increased mechanization of the fresh-produce industry could boost productivity, ultimately helping to tamp down price growth. It also could help farmers in California, who are struggling with a yearslong drought in the country’s largest produce state, get more from their fields, offsetting higher costs.

A machine developed by a Spanish entrepreneur automates the process of picking small produce like strawberries. Many farm owners hope it will help alleviate the impact of labor shortages. The labor shortage spurred Tanimura & Antle Fresh Foods Inc., one of the country’s largest vegetable farmers, last year to buy a Spanish startup called Plant Tape, whose system transplants vegetable seedlings from greenhouse to field using strips of biodegradable material fed through a tractor-pulled planting device.

On a recent morning in Salinas, Calif., Tanimura & Antle Chief Executive Rick Antle watched as two workers fed romaine-lettuce seedlings-—encased in the biodegradable strips like a belt of machine-gun bullets—into the device, which precisely cut the seedlings and drove them into the soil. The machine cruised at more than 6 miles an hour—lightning fast for a produce field. In commercial trials, Plant Tape has eliminated at least 10% to 15% of the overall work hours for growing romaine and celery, Mr. Antle said. Plant Tape is now ramping up production so more Tanimura & Antle fields can use it.

agrobot2

Meanwhile, on the same day, the old method was still on display at a Tanimura & Antle celery field down the road. Half of a team of 16 workers slipped celery seedlings into a tractor-pulled machine that plugged them inches apart into rows of soil. The other half, trailing behind, manually smoothed dirt and straightened any misaligned plants. The operation covered about nine-tenths of an acre per hour, which at its fastest was equivalent on average to 0.9 miles an hour.

Tanimura & Antle in recent years has resorted to bringing hundreds of workers in from Mexico on costly, temporary visas for such work. The decades-old system needs to be replaced because “we don’t have the unlimited labor supply we once did,” Mr. Antle said.

Machines are doing more than picking produce. Altman Specialty Plants Inc., one of the country’s largest nurseries, has been using eight, squat robots for the past two years to ferry more than 1.2 million potted roses and other plants to new rows as they grow larger. The $25,000, self-driving machines have occasionally gotten stuck in mud, but they freed eight workers for other jobs and ultimately paid for themselves in 18 months, said Becky Drumright, Altman’s marketing director.

“This is the least desirable job in the entire company,” she said. With machines, “there are no complaints whatsoever. The robots don’t have workers’ compensation, they don’t take breaks.”

‘There are no complaints whatsoever. The robots don’t have workers’ compensation, they don’t take breaks.’
said Becky Drumright, Altman’s marketing director Reiter, the Driscoll supplier, is starting to deploy joystick-guided and remote-controlled machines that carry building supplies and boxes of picked produce through berry fields. They free up hundreds of worker hours a year that can be used for other tasks, said Nathan Dorn, the company’s director of farming systems.

Meanwhile, Mr. Bravo is planning a new version of his Agrobot strawberry harvester that will have 60 robotic arms, and require strawberry production on raised, hydroponic beds instead of low, dirt fields. Workers will simply sit atop the machine inspecting and packing berries that pass on a conveyor belt. He plans to begin commercial field trials with his machine this fall.

Last week, I presented a new construction machine to replace masons.  This week, a robotic harvester to replace farm hands. As more technology takes over the most unskilled and vulnerable workers, the ethical demands on a robotic society will only grow if not addressed directly by industry leaders. For example, for every worker we replace with bolts, we can invest a % of every dollar spent on robotics towards the re-education of human laborers.

Thank you to Ilan Brat of The Wall Street Journal for his original reporting of Argobot.