Disruptive Improvements on the Forefront of Change

Tune in for a dialogue on the Robotics Evolution and Revolution with Dr. Henrik I. Christensen, Chancellor Chair of Robotic Techniques for Qualcomm, Professor of Pc Science at Division of Pc Science and Engineering at College of California San Diego, and Director of the Institute for Contextual Robotics.

Webinar Transcript:

Jeremie Capron:

My title is Jeremie Capron. I am the director of analysis right here at ROBO International. As a lot of you realize, we’re the creators of research-driven index portfolios which are designed to assist buyers seize the expansion offered by the expertise revolution of robotics, automation, and AI. And that is what we’ll be speaking about in the present day with a really particular visitor. I am actually thrilled to be joined in the present day by Professor Henrik Christensen. He’s the director of robotics at UC San Diego. And Henrik can also be the principle editor of the US Nationwide Robotics Technique and an advisor to corporations, governments, and buyers all over the world. Henry can also be a co-founder of ROBO International, and he is been serving to us steer our funding analysis efforts over the previous seven years. And for that, we’re very grateful. Henrik, thanks, and welcome.

Henrik Christensen:

Thanks.

Jeremie Capron:

So, in the present day we’ll focus on a few of the most fun areas of robotics, automation, AI, and the way buyers can capitalize on these developments. It is a essential matter to us, to buyers. Robotics and AI, we see as a set of general-purpose applied sciences that may be utilized to each business very very like the web or electrical energy originally of the twentieth century, and that is providing large funding alternatives. So we’re very lucky right here at ROBO world to be working with a few of the most outstanding thought leaders in robotics and AI; lecturers and entrepreneurs like Henrik and several other others you could see right here on a strategic advisory board. So we’re actually proud to mix a monetary and funding evaluation strategy with expertise analysis and experience that is coming from folks like Henrik. And in the present day there’s greater than $4 billion in funds monitoring our methods. Right here on the New York Inventory Trade, in addition to in Europe and in Asia.

And essentially the most notable might be ROBO. ROBO was the primary robotics and automation ETF. It launched round seven years in the past. It is returned greater than 50% over the 12 months to the top of the second quarter, and an annualized 22% over the previous 5 years. So, we additionally run THNQ. That is the factitious intelligence index, which we launched three years in the past, and Henrik has been instrumental in serving to us analysis and perceive the world of AI machine studying. And HTEC; H-T-E-C, that is the healthcare expertise and innovation index, and you’ll see the three methods have considerably outperformed world equities since inception. You’ll find extra data on our web site at roboglobal.com. So with that, I wish to invite Henrik to the digital stage. Give us a short presentation on robotics evolution and alternatives. And for all contributors, we’ll be addressing your questions after that. So please use the Q and A operate on the backside of your display to log your questions. All proper, Henrik.

Henrik Christensen:

Hey, thanks, Jeremie. Let me see if I can get the sharing going right here. There we go. So, thanks, Jeremie. Thanks for this chance to additionally speak about a few of the issues I am very enthusiastic about. So, one of many issues I have been doing, as Jeremie talked about, is that I am the principle editor of the US Nationwide Robotics Roadmap, which we publish each 4 years. So 2009, it was, after all, the financial downturn. It was collaborative robotics that was the principle stage. 2013, after the horrible catastrophe in Fukushima, it was robots for rescue and people sorts of issues. 2016, it was globalization. How will we then get excessive combine, low quantity? And now the roadmap that is simply been revealed, we’re taking a look at what’s the post-pandemic world that we’re taking a look at. So we have a look at a few of the mega developments.

It’s how will we get to a society the place we’re put up COVID? What are the impacts there? We’re seeing nonetheless altering commerce dynamics. We’re seeing mass customization, questions on urbanization; are all of us transferring again to the large cities, or are we going to remain away in kind of a work-from-home form of situation? And naturally, the ageing situation or ageing society remains to be a reasonably large deal. If we have a look at a few of the COVID implications, it’s totally clear that logistics’ last-mile deliveries, these sorts of facets, are taking part in an enormous function. It is grown tremendously. We had 18 months or one thing like that, the place most individuals would take house deliveries relatively than going to grocery shops. We noticed someplace between 40 and 60% progress in last-mile logistics. That can proceed. We see that staying on. However we additionally seeing Tele-robotics, the place there’s a whole lot of curiosity for medical doctors and others to not should see sufferers nose to nose, however nonetheless conducting digital visits.

We’re seeing, after all, screening. We’re nonetheless doing COVID testing on a really giant scale. We’re seeing meals security. How can we assure that some meals was not touched by people? Cleansing, disinfection. And naturally, we additionally seeing a scarcity of a migrant workforce. It was once that you’d have lots of people transferring round due to COVID and due to different facets, we do not have that very same dimension of a migrant workforce, and that impacts how we might do issues, and it motivates a better diploma of automation than we have seen earlier than. When it comes to the commerce dynamics, we’re seeing manufacturing transferring in areas like automotive. 30% of the automobiles are nonetheless made in China, however we’re seeing a motion in direction of India, in direction of Vietnam, in direction of different areas which are impacting this. Wage remains to be an enormous issue, and for that cause, a few of that manufacturing shouldn’t be transferring again to the US, and that motivates a better diploma of utilizing automation as we see a shuffle, identical to we did in 2009.

In order that’s necessary to bear in mind, however clearly, the US-China is impacting how we have a look at this for the long run. Some will come again to the US. Some will go elsewhere. Mass customization is and can proceed to be a significant pattern. All people desires a one-off. So if you happen to have a look at BMW X5 that you just’re seeing on this slide right here, it is out there in 4 million completely different configurations. All people desires their very own form of seats, their very own sorts of audio system, their very own completely different sorts of engines. All of that suggests that these automobiles can be found in 4 million completely different configurations. We solely make the identical automobile each two and a half week, which indicate that you have to have a extremely streamlined provide chain and manufacturing course of, and you’ll solely actually do that you probably have a really excessive diploma of automation.

We’re seeing the identical factor for particular person merchandise like sneakers. All people desires their very own coloration, their very own customization on it. 3D printing and different applied sciences are enabling us to do that to a level we have by no means seen earlier than, which is admittedly cool. My long-term imaginative and prescient remains to be that you will have an autonomous UPS truck driving round in your neighborhood, and it will likely be 3D printing a product for you. And so that you order it, 5 minutes later, we’ll ring on the door and we’ll say, right here you go. Getting older society remains to be taking part in into this put up COVID or pre COVID. We had eight hours of enhance within the age of the world inhabitants every single day. We have seen a setback by two years throughout COVID, so the general common age of the world inhabitants went down by two years.

That is beginning to decide up once more. So once more, you must count on to see eight hours a day. This suggests that we are going to stay longer. Sadly, we’ll get numerous sorts of challenges as we get older, each when it comes to mobility and when it comes to different issues that may are available and affect this. We not solely have societal drivers, however we even have expertise drivers. So, new supplies, for example, are essentially altering how we do that. We will construct grippers which are absolutely personalized to merchandise. So, for example, if you happen to do strawberry choosing, we are able to really construct a gripper that morning that can be utilized for the scale of strawberries we’ll decide in the present day. And tomorrow, it could be a special dimension for strawberries. We will actually customise it to this utilizing a mix of beta indicators, 3D supplies, and shortly we will even have a significantly better hybrid sensing constructing, which essentially modifications how we do that.

So we are able to do print to order, and we are able to add sensors that we have by no means seen earlier than. The identical factor with sensors, we have seen digicam expertise popping out. Right here, the grey space is conventional old-style cameras, blue is your normal journey cameras, Orange is your cellphone expertise. So, you are having extra of those come out that we produced every other form of digicam. I stay in San Diego. So Qualcomm would inform me you could purchase cameras for these in quantity at $1, and you should purchase the processing energy to do it for $9. So instantly we’re at a degree the place we are able to use applied sciences that we have by no means imagined earlier than, which is admittedly cool. We additionally seeing computing. Computing is turning into very low-cost, so we are able to now do trillions of operations per second, together with deep studying, and all of this, we are able to combine with eight to 10 cameras right into a single processor, and get that processing energy.

So we getting to a degree the place we have commoditized computing, AI, and deep studying to a degree that we’ve by no means imagined earlier than, which is admittedly cool. The truth that I can do all of this and put 10 bucks at quantity makes a distinction in how we take into consideration issues. Another societal drivers, if we go briefly by way of a few of the verticals, provides you with a way of the place we’re going. So in manufacturing, it is nonetheless value, it is nonetheless flexibility, nevertheless it’s additionally this that we are able to do very excessive quantity but in addition excessive combine in order that we are able to get nearly one-off manufacturing. Identical time we get into the place we’ve codeless programming. So you possibly can think about you possibly can roughly present up and you should utilize a quite simple interface to have the ability to do very simple programming of this.

It was once that it is required you to have weeks of coaching to have the ability to function a robotic. Right here, you are kind of seeing an iPad-like interface that makes it quite simple for folks to truly go in and do programming of this. So we’re seeing READY Robotic, we’re seeing Intrinsic, we’re seeing numerous these corporations which are popping out with quite simple interfaces that transform the price of deploying these techniques in on a regular basis situations. Logistics pushed by Alibaba, Baidu, Amazon, or Walmart, but in addition pushed by the truth that e-commerce is admittedly driving this ahead. You will note far more versatile distribution facilities than you have ever seen earlier than that permit us to go in, put materials and actually transport it to you in a minimal period of time.

When you go and ask UPS, UPS will say, once they ship the bundle to you, three hours of the time, it sits nonetheless, the remaining quantity of the time, it’s really transferring in direction of your vacation spot. So the truth that we getting this degree of agility makes an enormous distinction for us. And particularly as we see a change in a post-COVID world in direction of far more e-commerce, far more supply of meals, no matter supplies, these are going to be the enabling applied sciences to make that attainable. Healthcare, we additionally see, the place there’s a whole lot of curiosity in individualizing it, ensuring we are able to do it. How can we ensure the sufferers, the households have entry, the combination? Right here, we’ve kind of the interblade between robotics, AI and healthcare applied sciences that we’re seeing in ROBO when it comes to how we are able to do that. As a result of we see an ageing inhabitants, this isn’t solely going to be what you see within the hospitals, nevertheless it’s additionally going to be, what are you able to do within the house? How will you do rehabilitation? How will you assist folks which are remoted to really feel much less remoted than they really feel in the present day?

So this is an instance. When COVID began, numerous medical doctors confirmed up and stated, I do not need to be proper subsequent to a affected person as a result of I am apprehensive about potential COVID an infection. So having a teleoperated robotic the place you possibly can stand even 10 toes away and do a teleconsultation together with your affected person would make an enormous distinction. So we have seen a elementary change in expertise is right here to offer distant entry, and distant entry might be wherever from 10 toes to lots of of miles, the place you present that degree of interface. That is in place now. We’re seeing the identical factor for having the ability to do distant surgical procedures. We have achieved experiments lately the place you will be on an plane service in the midst of the Pacific, and a surgeon can dial in and help with doing surgical procedure for this with out having to truly be on the plane service.

So we’re revolutionizing healthcare and entry to healthcare at a degree we’ve not likely seen earlier than. We’re seeing new sorts of mechanisms. Digit is a robotic that was developed in Oregon that… So the issue is with a whole lot of the last-mile deliveries, you possibly can solely get to the staircase. We’re seeing innovation in new mechanisms that may stroll up the staircase that may get you all the way in which to the entrance door. It is a expertise that’s being examined by a number of corporations proper now to truly make this attainable. It is not solely a query of, can I make it attainable, nevertheless it’s additionally a query of, can I make it accessible when it comes to financial prices? And right here, we’re seeing these robots popping out and doing this in a means that could be very, very versatile. So you will note these new applied sciences.

One of many areas the place we have been lacking is when it comes to greedy and manipulation. It has been very onerous, if you consider it, at your fingertips, you’ve got about 300 mechanoreceptors that can help you have very agile management. Till lately, that was troublesome to do. With new supplies, we are able to now really go in and construct very built-in gripper expertise that may permit us to do manipulation as we have by no means achieved earlier than. That means as I discussed early on, we are able to do the choosing of strawberries. We will deal with grapes in vineyards. So it is all of those that is popping out, and on the similar time, we come at it at a value that’s accessible to even comparatively small-sized corporations. When it comes to sensing and notion, we have seen the evolution of AI and machine studying that has enabled us to get a lot, a lot additional than we have achieved earlier than.

We will use this, as you see within the high picture, to go in and acknowledge the entire issues in a front room or in no matter space. So, whether or not you are constructing a wise vacuum cleaner that is aware of, oh, wait a minute, I am going beneath the sofa, that is the place all of the mud is, we are able to construct in that form of information, however we are able to additionally use it to do autonomous driving automobiles in order that we are able to drive coast to coast with out driver intervention. The identical factor, we are able to do monitoring of individuals, so we are able to do monitoring of them. We will do biometrics on this. So it is fascinating that we have seen this very low-cost digicam expertise, very low-cost computing energy with a mix of AI that is going to revolutionize how we construct clever techniques for the long run. We have additionally seen kind of it was once that it was very onerous to do programming of those techniques in an inexpensive means.

We are actually attending to codeless programming the place you do not have to write down code in a standard vogue, as an alternative, you possibly can go in and transfer round just a few blocks. And by doing this, you possibly can fully reprogram a producing line or a provide chain line to do that. That is very fascinating as a result of it allows us to construct a paradigm the place you want very restricted coaching and you’ll nonetheless get to an excellent place of doing this. So, it was once that our rule of thumb was that for a system being deployed, 25% of the fee was the robotic, 25% of the fee was for the {hardware}, after which half of the fee was really software program. If we are able to take that fifty% of the fee and cut back it to one thing like 10%, the price of the general system goes down by 40%, and that opens up numerous markets we have by no means seen earlier than. But it surely additionally implies that we are able to have folks which are largely unskilled labor that may design and program these techniques.

So that is actually altering how we enter the market and why we’ve seen such an amazing quantity of progress that we have not seen earlier than. In fact, I’d be amiss if I additionally did not point out that AI, machine studying, and deep studying is revolutionizing wherever from automating electrical processes, however all the way in which right down to doing logistics, to doing manufacturing. We will do significantly better analytics. We will individualize it. We will arrange these techniques. We will program them to have a really excessive diploma of agility, which we have not actually seen earlier than. So that is an space the place we’ll proceed to see numerous vital progress. It was once that it was primarily a few of the large, the FAANG corporations that have been doing this, however now we’re seeing this getting commoditized right down to a degree the place all people is utilizing it.

Identical time we’re getting to a degree the place we’re utilizing standardized or kind of pre-trained techniques to do that, so you do not have to spend days doing coaching. We will discuss for about minutes or seconds and nonetheless get very excessive efficiency. So in abstract, the world is admittedly adopting robots. We’re doing this to extend the standard of life. So we have seen these folks need to stay autonomous. All through this, we would like to have the ability to assist with healthcare expertise to ensure we are able to tackle regardless of the challenges are. We enhance the pliability in order that we are able to get to the e-commerce, we are able to get to tens of millions of various configurations of assorted merchandise, and on the similar time, allows us then to make use of this to do meals, addressing local weather challenges, addressing financial progress, and democratizing it, so it turns into out there for everyone.

And that is why we proceed to see total progress for the complete sector of 15 to 18% when it comes to adoption, when it comes to business statistics, and that then, in flip, we see very vital progress for numerous completely different corporations. And with that, I’ll hand it again to Jeremie.

Jeremie Capron:

Thanks very a lot, Henrik. Let me ask you just a few questions whereas we await the contributors’ questions. I will remind all people, you should utilize the Q and A field on the backside. I see we’ve just a few already coming in, however let me kick it off with… Properly, you are working with a whole lot of college students, entrepreneurs, governments, large companies all over the world. I do know you are simply again from Europe. Let me ask you this, what’s essentially the most thrilling challenge that you just’re concerned with in the present day or within the current previous?

Henrik Christensen:

So for me, essentially the most thrilling is that we constructed autos for doing micro-mobility, the place we are able to do bundle deliveries in very crowded environments. So the truth that I might drive one among my autonomous driving autos on fifth avenue in New York, and in a position to really handle to undergo the entire pedestrians and do that in a means that’s acceptable. The truth that we have gotten to a degree the place I could make this low-cost sufficient, I could make it accessible sufficient, and I can do it in a means the place folks can drive by way of it, I believe, is credibly fascinating. In order that’s kind of one. So doing the micro-mobility. I believe the opposite one was the challenge that I discussed at our analysis institute we now have in order that we are able to do teleoperation, teleoperated surgical procedure on an plane service in the midst of the Pacific. This opens up very fascinating alternatives to do that. So these are two examples the place I really feel we’re actually altering the world.

Jeremie Capron:

Proper. Yeah. Henrik, we’re very excited concerning the deployment of autonomous techniques. We have seen them in manufacturing. Now we’re seeing them in logistics with autonomous cellular robots, transferring items and elements in warehouses, and now you are speaking about micro-mobility and precise autos on the street in public areas. We have seen robo taxis going business with values levels of success, nevertheless it’s right here. In reality, we’re additionally taking a look at an organization that makes autonomous indoor drones that may fly in swarms. For logistics functions, what do you assume is the present cutting-edge for autonomous techniques, and the place do you see essentially the most business attraction?

Henrik Christensen:

I believe I will break it into two as a result of there’s the indoor and there is the out of doors. And for the out of doors, we are actually beginning to see coast to coast autonomous driving with these autos, the place you get on the freeway. So it’s kind of a ramp to ramp autonomy. Numerous that is really not pushed by, can I get the motive force out of the cabin. It is pushed by that we see 5 to 10% gasoline financial savings. So automobiles are literally higher at driving, or computer systems are higher at driving than persons are. I can do coasting. I can predict what the prepare appears like. I can predict the site visitors. And if you happen to have a look at this from, if you’re a few of the large carriers, 5-10% gasoline financial savings is an amazing sum of money. So there, I believe we’re seeing this, and I believe we’ll see a whole lot of them, the expertise being matured on the interstate for ramp to ramp autonomy earlier than we actually get into the cities. It is extremely onerous to do driving in the midst of this.

So I am apprehensive about when sure corporations promise we may have degree 5 autonomous automobiles and we may have robo taxis. We’re not fairly there but. It’s going to take somewhat little bit of time. In order that’s why I believe the interstate might be the place we’re doing the testing. Indoor for autonomy, we’ll see absolutely autonomous warehouses which are doing this, but in addition we’ll see as you talked about, locations the place you utilize drones, for example, to do stock administration, the place you possibly can go in and you’ll monitor the entire issues which are in your warehouse utilizing drones as a result of it is a lot simpler than going and utilizing a staircase to climb up and down the cabinets to determine what’s on the cabinets. So I believe we’ll see indoor touchless much less monitoring and administration of warehouses, and out of doors, it should be ramp to ramp autonomy that is going to drive the economic system.

Jeremie Capron:

Okay. Properly, I see we’ve a few questions across the affect of the pandemic, and I believe clearly the pandemic has turbocharged the digitization of the economic system. It grew to become very obvious final yr. Most corporations we analysis and discuss to the members of the ROBO index. They inform us that automation demand is elevated considerably, and now the monetary markets definitely replicate that to some extent, proper? The ROBO index portfolio was up 45% final yr. The AI portfolio was up within the sixties, they usually proceed robust into 2021. So to me, that is considerably much like 2017. The final time we have been listening to from automation corporations that demand was extraordinarily robust. What do you assume is occurring right here and the way sustainable is that new trajectory?

Henrik Christensen:

Properly, so I believe the 2 completely different questions in there. So when it comes to the post-COVID logistics automation, I believe folks have realized, why do I must go to the grocery retailer for lots of those on a regular basis purchases. I can go in, I can get it delivered. In case you are a mother or father with two children that you just in any other case must decide up in daycare, undergo the grocery retailer, the place you are attempting to do crowd administration of your loved ones when you’re additionally attempting to do that is extremely demanding. You notice I haven’t got to do that anymore. I can go house, I can order it. And any person will present up at my doorstep, and that essentially modifications it. So I believe we have definitely if I have a look at my family, we’re in a degree the place we go to the grocery retailer lower than half of what we did earlier than since you simply go in and you place it on the record of what do we’d like.

After which finally on the finish of the day, you press please ship tomorrow earlier than midday, and it simply occurs. This provides us a whole lot of flexibility that we did not have earlier than. I haven’t got to fret about which retailer do I’ve to go to. It will get delivered. So I believe that is right here to remain. It offers us a degree of flexibility we have by no means imagined earlier than. I do not see folks going again anytime quickly. There are nonetheless those that need to go in and say, no, I would like the most recent, I would like the freshest greens and stuff like that. That’s 10 to fifteen% of the shopper base. It is not the bulk.

So the bulk will do that. And the truth that you will get it to your entrance door makes an amazing distinction. When it comes to the digitization of the economic system, which was the opposite a part of the query, is that we are actually at a degree the place, initially, due to sensing, we are able to observe each product. All the way in which, I can inform the place it’s from the time the strawberries have been picked in San Luis Obispo till they get delivered to my entrance door.

This allows me to do significantly better analytics. I perceive the place’s the loss. The place is it? I can observe it. Is it recent? How recent is it? So all of this, and the identical factor for the provision chain administration. So the instance I discussed earlier than, in an ordinary automotive manufacturing unit in the present day, storage capability is 12 minutes. Stuff arrives on the manufacturing unit, 12 minutes later, it is mounted within the automobile and it is on its means out. So we have gotten to this the place we used to have large warehouses, a whole lot of storage capability constructed into the system. These days are gone. We will use AI, we are able to use analytics, we are able to do predictive management to foretell, what do we’d like for Thanksgiving? What do we’d like for Valentines?

And all of this enables us to do significantly better administration of our capability. In fact, then the problem is we have made these techniques extremely lean. So if you happen to see a problem just like the Suez Channel, instantly our system is, uh-oh, we did not construct in storage capability to deal with this as a result of… So our system remains to be not as strong as you desire to, however total, I believe we made these techniques extremely lean by way of AI, by way of automation. It’ll stick with us, however we nonetheless have to determine how will we robustify these techniques for large challenges just like the Suez Channel.

Jeremie Capron:

Henrik, we’ve some questions across the labor affect and there is a labor scarcity proper now within the U S definitely popping out of the pandemic. The place do you see robots serving to remedy that drawback, and longer-term, how do you see the affect on the workforce?

Henrik Christensen:

So, we did kind of an evaluation for the roadmap the place we checked out this, and there is nearly one correlation between investments in robotics and progress in jobs, and it isn’t causation. It is not like if you happen to spend money on robotics, you additionally make use of extra folks, however there is a robust correlation which has to do with stronger economies. So I believe it will likely be the soiled uninteresting and harmful jobs which are getting changed my labor. When you have a look at the usual warehouse, the turnover is someplace round three folks a yr. Individuals do not need to work in these warehouses. There is not any means you see, it is a hundred levels, the humidity is 90. This isn’t the place folks need to be. So we’ll see numerous these jobs will get changed. And on the similar time, it’ll give us a lot increased agility when it comes to managing these processes.

I can construct warehouses in my native neighborhood. So it is a whole lot of these course of that we are going to see, the place we’ll get in. And the goodness is that permits us to displace folks to the areas the place you really want the human contact. I am nonetheless shocked that in the present day… So the opposite factor I do, I am a meals connoisseur. I like to go to eating places and stuff like that. Many eating places are nonetheless solely at 50% capability as a result of they will get folks. So, we have to go away from the locations the place we actually do not want folks and put them in locations the place we actually want folks.

And so we’ll see a shift in workforce round this. I am not seeing at the least within the brief time period, a major discount within the want for workforce, it should proceed to develop, however it will be the very conventional, extremely repetitive jobs in powerful work environments that may go away. That is the place we do the alternative. That is true for India. It is true for China. It is true for the US. So it is a world effort that we are going to see. However we have achieved many research and up to now, we have not been in a position to present that at a macro scale, robots are changing jobs. It is shifting jobs to new domains, nevertheless it’s not eliminating jobs.

Jeremie Capron:

Understood. I see some questions round AI coming in. As you realize, Henrik, we predict AI is admittedly a type of key enabling applied sciences which are increasing the scope of potential functions for robotics. And it is actually an necessary a part of our funding technique, each for the ROBO and financial institution portfolios. So, lots has been achieved round pc imaginative and prescient and the popularity round that in digital footage after which the pure language processing we have made large progress. Now, AI is pushing into different broader ranges of functions. The place do you see AI being deployed efficiently? The place is it going? I see, we’ve a query concerning the 5, to 10-year horizon. Perhaps we needn’t go that far out, however the place is it going?

Henrik Christensen:

So for me, it is, we have seen an amazing quantity of progress on… a lot of the progress we have seen on AI has been offline. The truth that I can do face recognition in my picture library, the truth that I can do a whole lot of… nevertheless it’s largely been the place I might do a whole lot of coaching forward of time. We’re now transferring to a spot the place I can do a whole lot of coaching on the fly, and doing this so we are able to actually get to real-time AI. And that might indicate that it will assist me take higher footage on my cellphone. So if you happen to take… I used to be very impressed just a few years in the past, I went to Huawei’s manufacturing unit in Shenzhen, and I stated, that is a whole lot of processing energy. After which it is like, oh yeah, everytime you take a portrait together with your digicam, we’ll take 60 photos.

And we’ll solely return the one the place you smile. So, for example, it is a digicam that takes superb footage. Yeah. As a result of we use AI to throw away the 59 footage that we do not assume you desire to. So that is an instance the place we’re getting from picture processing, from doing this… I flew again to, as you stated, from Europe lately, and Atlanta airport, they do face recognition of you. You do not have to indicate your ticket anymore. That is the place we’re getting all of this. The consumer interface will essentially change how we do that. In order that’s one space of AI, we’ll do significantly better analytics than we have achieved earlier than. After which, after all, we’re attending to the place we are able to do pure language processing. So the place we are able to do translation, we are able to allow it.

Now, I used to activate grammar checks, blah, blah, blah, in phrases, to make it possible for I really wrote an inexpensive set of texts in the present day. I’ve all of those AI instruments that stated, oh, you’re conscious, you must have a remark right here, and you must actually change this to this different phrase. So when it comes to NLP, we’ll all develop into prolific, superb writers. Regardless that naturally I won’t be a prolific, superb author. So we’re seeing this, it is enabled us in our day by day lives to compensate for our limitations and bringing them ahead. So I believe that is good.  The third space is in healthcare. We will do significantly better diagnostics. We will monitor you. The scary half can also be, I’ve all of those applied sciences that monitor me each day that may inform me, are you figuring out onerous sufficient?

Are you exerting your self an excessive amount of? Are you conscious your temperature simply went up 0.5 levels? Perhaps you must take it simple in the present day. So we’ll get this huge quantity of knowledge, and we have already had that huge quantity of knowledge, however now we are able to course of it and supply choice help to folks. Whether or not it is on your day by day train routine, it is for meals it is for your corporation processes. I’d say we have solely touched the primary 5% of AI. So that is an space the place you are going to see huge progress going ahead. The identical for TV suggestions, no one sits down and does the schedule and says, what’s on at eight o’clock? You go and say, hey I need to watch a comedy proper now or…

So we’re attending to the place we’re far more personalised in something we do, and it is being enabled by AI. And it is as a result of it is turning into so low-cost. As I stated with my digicam, I can now get to the place I can do that for 10 bucks. And that suggests that you are able to do it in all places on your housebreaking system, on your TV leisure, on your meals prep, that modifications it. So I believe that is why we’re very excited concerning the THNQ portfolio as a result of it opens up for applied sciences that may affect all facets of your day by day life.

Jeremie Capron:

And following up on that, Henrik, I see some questions across the software program facet, the {hardware} facet of the robotics business. What are the enabling applied sciences that matter essentially the most to the continued enlargement of robotics? We have now a participant asking if it is extra concerning the software program and the algorithms or the machine itself and the elements. I am curious to listen to, the place do you see the alternatives when it comes to enabling applied sciences?

Henrik Christen…: Certain. I have been quoted earlier than when it comes to saying that I really feel that the final 50 years of robotics have been dominated by mechanical engineers. It has been about constructing mechanisms, constructing them so that you’ve excessive precision, you’ve got a excessive raise, you’ve got excessive effectivity. All of these items are actually within the software program. And I believe the enabling half has been, we have got two sensors which are kind of strong sufficient and low-cost sufficient that we are able to instrument every little thing. On the similar time, we would gotten to the quantity of processing energy, and there it is each the CPU, but in addition gaining access to the graphical processing models, the GPUs that may do huge quantities of knowledge processing in very restricted time. So these instruments have now been commoditized and made sufficiently small you could put them in all places. And now it is the software program. On the similar time with the software program, it was once extremely difficult to program these techniques.

Now we’re attending to the place by way of machine studying, we are able to do data-driven extraction of key parameters. So if you happen to have a look at a few of the AI-based corporations, they are going to do pre-training so that you just’re 90% at an answer earlier than you present up on the buyer. And you then present up on the buyer and also you do the final changes, and you might do that, it was once days or even weeks. Now you are able to do this in seconds and minutes. Come and alter to the actual line. So we’re seeing this, so I am positive a few of you noticed that Google simply spun out intrinsic, which is attempting to go after a means of truly doing this. We have seen co-variant, we have seen corporations like Berkshire Gray which are attempting to do these items. So numerous these corporations which are popping out that can help you very quickly arise a system and get it going.

And so we’ve now the sensors, we’ve the computing energy, but in addition we’ve software program instruments that make it accessible to non-PhD individuals who really do programming of those techniques and doing it a really restricted period of time. So earlier than I stated 25, 25, 50, when it comes to value, that fifty% software program value has come right down to one thing that’s a lot decrease and with a a lot increased diploma of agility. So, that essentially modifications it. Nonetheless, we’d like higher instruments. However the different factor is that we’re getting very large-scale information units that permit us to do a whole lot of the coaching earlier than. Earlier than it was onerous. Any individual got here to me and stated, we’ll use a petabyte of knowledge to coach your system. I’d have been, that is actually scary. At present, put them on video, positive, not an enormous deal. However it’ll take us just a few hours to course of it, however 10 years in the past that might have been intractable.

So the query is, how are we… the subsequent factor is, I believe the place we’ll see the subsequent revolution is in hardware-software co-design. The place we’ll get to a degree the place you present me the general wants of your software. And I optimize it for what ought to I do in {hardware}? What ought to it do in software program? And so 5 years out, {hardware}, software program co-design, and that is the place a few of the corporations we have seen much less progress from. So AMD, Intel, all of those corporations will come again as a result of this can open up solely new methods. We have seen progress that we have not seen lately.

Jeremie Capron:

I believe there’s an necessary query right here across the environmental impacts and the way these applied sciences can maybe assist. And right here at ROBO International, it is a essential matter to us. Not solely environmental however social and governance facets. We pay a whole lot of consideration to these elements after we design these portfolios. I believe on the whole, there is a pure tendency for robotics and automation to have a constructive affect when it comes to the surroundings, simply given that they are targeted on productiveness and effectivity and utilizing much less vitality and pure useful resource. However then again, you even have this dramatic enhance in compute energy and we’re utilizing increasingly chips and maybe some several types of supplies that might be more durable to recycle. So we’ve a query from a participant about AI serving to with local weather change. What are your ideas round that, Henrik?

Henrik Christensen:

So I believe, there’s fairly numerous completely different facets to it. So I discussed earlier than, driving coast to coast in vehicles, if you happen to do autonomous driving, we are able to cut back the gasoline consumption by 5% %. That is a reasonably large deal that we are able to really go in and do that. Computer systems are significantly better at anticipating site visitors, anticipating when can I do coasting? When can I not do coasting? In order that’s an instance the place we see it, but in addition we’re seeing it when it comes to having the ability to do manufacturing on demand. So it isn’t like if you happen to’re harvesting fruits or no matter, that you just’re apprehensive about, am I really constructing stuff or producing stuff that isn’t wanted, which might affect when it comes to, I stay in California, so I care deeply about water. So having the ability to do that.

So I believe we’ll get to a lean manufacturing system the place we save water. We solely apply pesticides the place we actually want pesticides. So all of this can permit us to construct a way more sustainable economic system. The draw back has been, for example for the deep studying, that a few of the very giant scale, for example, pure language processors takes extra CPU energy to coach than 100 flights throughout the Atlantic. In order that’s fairly scary that a few of these coaching… you are taking a super-computing middle at one of many large corporations, and also you run it for 3 days. That generates as a lot CO2 output as flying throughout the Atlantic 100 instances. So there we are actually attending to the place we will be a lot smarter about how we do coaching, but in addition we are able to stage it so we are able to do pre-training, which can be utilized time and again.

So we used to do group power functions of computing energy, however we’ve now acknowledged that until we’re very cautious about how we use these techniques, this might have a destructive environmental affect. So if you happen to go and ask the large corporations, the primary query is, what is that this going to do for the local weather? The identical factor we’ll see for all of the issues I believe we are able to optimize processes to be far more environmentally pleasant than what we have seen earlier than as a result of we are able to actually make them lean. We will actually make them optimize whether or not it is driving or manufacturing. We will optimize it for supplies to do significantly better recycling than we have achieved earlier than. We will kind recycling significantly better than we have achieved earlier than.

After which lastly, we’re attending to the place the place we apply AI in a stage vogue to make it possible for it doesn’t have a destructive affect on the surroundings. So I believe you must count on to see this far and wide, but in addition you must count on that that is one thing we see far more emphasis on. 10 years in the past folks have been like, yeah, in a second. Now, that is query primary or quantity two, after we speak about how will we consider processes?

Jeremie Capron: 

So, we’re operating somewhat bit out of time, however I need to tackle a pair extra of the questions coming in Henrik, particularly round China. Right here at ROBO, we have chosen to incorporate in our portfolios, some expertise and market leaders in China. And China has been the fastest-growing marketplace for automation for a while now and lately obtained a few of the native champions displace overseas leaders. What do you consider the US-China expertise race, and maybe a few of the implications we should always take into account as buyers?

Henrik Christensen:

Properly, so I will offer you a number of solutions to it. So the primary is, after all, as ROBO world, we’ve not solely a US funding automobile, we’ve, the index is Hong Kong, Singapore, Australia, London, you title it, we’re far and wide. So for us, we are able to’t afford to solely take into consideration this from a US perspective. We have now to consider this from a worldwide perspective. The fastest-growing marketplace for AI or for robotics in the present day is China. So I discussed earlier than 30% of all automobiles manufactured worldwide are made in China. And most of the people won’t notice this. The US shouldn’t be even within the high 5 for manufacturing automobiles. It is China, it is Japan, it is India, it is Germany and it is Korea. After which quantity six is the US. So we have to be cautious. And in automobiles, particularly, there is a large adoption of robotics expertise to make this attainable.

So we see yr over yr continued 50% plus progress in China. And we’re seeing, Midea acquired KUKA. They’re the third-largest robotics firm on this planet. That’s nonetheless going to be an important half. We are going to see a few of that transfer out of China due to the US… China’s different commerce complicated, or we’ll see a few of it transfer out. I am not satisfied it should transfer again to the US. I believe it’ll transfer to Vietnam. It’ll transfer to India. India simply introduced this large initiative of, make it in India. The home market in India is as giant because the Chinese language market. It is not the US. So we’ll see a diffusion of expertise going over this. However on the similar time, we’ve to remember that with these very giant home markets, we’ll see expertise.

Additionally in AI, we’re seeing Baidu, we’re seeing Alibaba. We’re seeing numerous these corporations which are very robust world gamers on this market. And I believe on the finish of the day, our accountability to our buyers is to make it possible for we return most capital to them. We need to have this and we are able to’t afford to be, at the least from my perspective, caught up an excessive amount of within the political local weather between the US and China. We have now to take a look at this and we’ll proceed to see this. So in China, the salaries have gone up 350% over the past 10 years, and that has pushed for a a lot bigger adoption of applied sciences. It’s now dearer to do manufacturing in China than it’s to do in Mexico. Simply to offer you a knowledge level, however nonetheless 30% of the automotive market.

We see different issues, textiles, all of those areas are in China. The provision chain is in China. For that cause, they’re automating. They’re doing the info science that we have talked about to a level that we have not seen wherever else. This is a chance. That is why you see ABB robotics headquarters in China, KUKA Midea is there. We’re seeing CSR, we’re seeing numerous these corporations which are doing investments there as a result of it is a very fascinating market. However I believe you’ll slowly see a diffusion. We are going to see it decide up in India, we have seen it already in Vietnam and in different areas. So we’ve to consider this as a global panorama and see, the place are the expansion alternatives? However you must count on to see precision gears, motors, CPU energy, all of that may have an amazing quantity of progress in China for the subsequent 5 to 10 years, at the least. And we have to at the least regulate this to ensure we’re making the precise funding selections.

Jeremie Capron:

Proper. Properly, as I stated, we do embody the Chinese language expertise and market leaders in our portfolios. And up to now it has been a great place to be regardless of the current considerations across the web giants in China getting beneath much more scrutiny from the Chinese language authorities that, fortunately that hasn’t actually affected the robotics and automation corporations over there.

Henrik Christensen:

On the similar time, the perfect autonomous automobile I’ve pushed, I drove in China. So Huawei took me on a tour exhibiting their autonomous driving automobiles exterior of Shenzhen, the place they have been utilizing a mix of sensors, CPU, and 5G. And it was simply unbelievable. I have not seen a demo like this wherever else on this planet. So, we have to regulate this to make it possible for we actually perceive who’s going to win this race. And I believe ignoring China could be naive. We have to regulate the place are these applied sciences and who do we predict are going to be the long run expertise leaders.

Jeremie Capron:

Okay. Properly, I believe it is time to wrap up right here. I need to remind everybody that if you wish to be taught extra about investing in robotics, automation, and AI, please go to our web site, roboglobal.com, the place we share a few of our analysis on corporations within the ROBO, THNQ, and HTEC portfolios. You’ll be able to join our publication. We’ll offer you a biweekly e-mail with a few of the analysis highlights. Thanks very a lot, Henrik, for sharing with us in the present day. I do know you are on a fairly tight schedule. I need to thank additionally all of you who joined us on this name in the present day, and we sit up for talking with you once more very quickly. Thanks all goodbye.

Henrik Christensen:

Thanks Jeremie.