As of September of this year, I have been actively analyzing the robotics industry for a decade.
During that time the industry has progressed greatly?products have come to market and proven themselves, companies have emerged and some have gone public, other companies have been purchased by larger organizations.
Robotics investment is up, as are installations. Robotics, it seems, has finally arrived.
With robotics, the exceptional has given way to the everyday. News concerning drone strikes and off-world rovers are commonplace. On two occasions, and in two locations, I have seen highway billboards advertising robotic surgery services. Articles on the Uncanny Valley run side-by-side with photos of the Kardashians.
Yet even after considerable mainstreaming, the robotics industry still boasts of much vigor and fresh faced optimism. Unfortunately, platitudes also abound. The revolution is just around the corner. Robotics is the first new industry of the 21st century. It is like the automobile industry in 1920s. It is like the PC industry in the 1970s. And on and on. It gets to be a bit much, but the passion and earnestness is genuine and hard not to like.
I myself have uttered these same such phrases, and frankly they do contain much that is true. Today, however, I am a bit more cautious. The robotics revolution, I have found, is not just around the corner. The revolution is always just around the corner.
No Place for Business Speak
In addition to enthusiasm and earnestness, the robotics industry (industries really) is unique in other ways. For example, the industry seems to be immune from the influx of cliched business speak.
I have yet to attend a robotics or automation conference and hear talk of dialed-in change agents opening up their kimonos with other thought leaders in their value chain to ballpark actionable ideas to incentivize their customer base and monetize their product lines. This is a good thing.
ISO 9000 and Six Sigma practices notwithstanding, the industry has also been largely untouched by management fads and fashions. You never hear ?robotics? and ?business process engineering? in the same sentence.
The same holds for ?Total Quality Management?, ?One Minute Management?, ?Management By Walking Around? and other management fads that have come and gone. Among my robotics associates, I know of no one who has Tom Peters, Peter Drucker or Jack Welsh on their bookshelves. Malcolm Gladwell?s “The Tipping Point” and Geoffrey Moore?s “Crossing the Chasm” are the rare exceptions that prove the rule.
Spoke Too Early?
Recently, a new business management technique ? the Lean Start-up methodology – has found favor with some in the robotics industry. Lean Start-up, and I am grossly generalizing here, is a methodology that reduces the chances of miscalculating and building a product no one cares about, or inversely, increasing the probability a developing a product customers actually want. It eschews traditional paths to business and product development involving lengthy planning and forecasting sessions, with the aim of releasing a full-featured product.
Instead, the Lean Start-up methodology calls for releasing a “minimum viable product” to early adopters that can be quickly validated. The product is adjusted as necessary (a ?pivot?) based on customer feedback, rapidly released again, and this process is iterated until it is determined that a ?product market fit? has been achieved. It is only then that the business side of the business, such as sales and marketing, is scaled up.
In many respects, Lean Start-up has transcended its role as a management methodology and is now a full-blown movement. This is especially true in Silicon Valley which, of course, is littered with high burn rate start-ups that crashed chasing mind share and first-to-market advantage with product concepts that found no resonance with customers. Lean Start-up was developed to overcome just this problem.
The Lean Start-up methodology does share many of the characteristics of other popular management techniques, both fad and non-fad versions. The technique?s developer and greatest proponent is the charismatic Eric Ries. Ries, an entrepreneur and adviser to technology companies and investment groups, introduced the methodology to the world via his widely read blog, Start-up Lessons Learned.
Ries’ product/market alignment technique found a ready ear among tech companies tired of wasting cash and cycles, and often whole companies, developing new products. In addition to increasing the odds of developing products the market will accept, the methodology is intuitive and relatively easy to implement.
As popular business and management techniques emerge, they invariably lead with a book and Lean Start-up is no exception. Ries?, “The Lean Start-up: How Today’s Entrepreneurs Use Continuous Innovation to Create Radically Successful Businesses”, debuted at #2 on the New York Times bestseller list. The title is a mouthful and a bit overblown, but who wants to merely create a successful business when a radically successful business can be had?
In keeping with other business management fashions, Lean Start-up has developed its own ecosystem of blogs (here, here and here), articles, and books (here, here and here), as well as a variety of conferences, workshops, meetups and seminars (here, here, here and here). Lean consultants too have emerged?and more are sure to follow.
Praise and Push back
Lean Start-up has many adherents, and has fostered much debate. But as with other emergent business and management methodologies, it was not long before push back began. Critics note that Lean Start-up is little more than a distillate of Steve Blank?s Customer Development Model (CDM), Dave McClure?s Acquisition, Activation, Retention, Referral, Revenue model (?AARRR? or Pirate Metrics), Agile Programming and other techniques and methodologies. Others have been less kind, describing what passes as Lean Start-up sagacity as trivial, banal, hokum and worse. While listening to one of the many Lean Start-up videos on the Web, I learned: ?It is not how much money you spend, it is what you spend money on.? Wow.
Critics also point out the few Lean Start-up use cases and success stories exist, and those that do, such as Dropbox, are not strictly lean (to be fair, the Lean Start-up methodology is relatively new). Also, many ?fat? start-ups, those that bring on loads of funding in an effort to grow quickly and dominate their market, have done well. Examples include Twitter, LinkedIn and Facebook, or if you go back a bit further in time, Google and Amazon.
A second and more philosophical Lean Start-up debate is focused on the process of innovation itself. Ries describes Lean Start-up as a scientific approach to entrepreneurship. That is, assumptions regarding functionality are testable and those that are proven true are incorporated into the next iteration of the product. This plodding approach contrasts sharply with a single-minded drive to produce a new product following an eureka moment of inspiration.
For many the debate comes down to customer validation (and cash preservation) versus vision (and higher risk). This is a false choice of course, as all start-ups (and mature established companies) employ each approach to some degree.
The Lean Start-up debate will continue. For the readership of Robotics Business Review, however, it is not whether Lean Start-up is effective, but whether Ries? principles can be applied to robotics. While any methodology can be adjusted to suit its purpose, the answer for the most part is a qualified ?no?.
Ries and other Lean Startup proponents have commented that the methodology works best for software-based businesses, and would be less effective for more hardware-centric products. I would put myself in that camp. On Ries? Lean Start-up website the case studies page only lists software firms ? Dropbox, IMVU, Wealthfront, Grockit, Votizen and Aardvark.
Lean Start-up practices are applicable not only to start-ups. They can be utilized by mature companies to bring new functionality or products to market. For example, Groupon, Zappos, Etsy, Heroku, Twitter and Zynga are all said to be employing Lean Start-up methods for their ongoing development efforts. It is important to note, however, that these firms are largely Web start-ups. Missing are Global 25 software firms such as Microsoft, Oracle, and Computer Associates.
You can be sure that developers at the largest software firms are aware of the lean movement, and evaluations are ongoing. But enterprise software systems, both front office and back office, are vastly different from the offerings of those companies publicly espousing Lean Start-up benefits (Heroku is the exception). It is not a matter of Web centricity or even mobile accessibility, but one of criticality, typically business criticality. In many ways the comparison is one of apps to applications. For enterprise software systems, the Lean Start-up?s build-measure-learn cycle can appear more like ready-fire-aim.
The Virtual and the Real
The term ?criticality?, returns us to the subject of robotics. For this discussion, a critical system is one where a lack of functionality would have serious, detrimental effects. Not all robotic systems are critical, but those systems in the fastest growing and impactful sectors ? industrial robotics, military robotics, autonomous vehicles, surgical and rehabilitation systems, and others ? are critical by any measure, including the ability to do physical harm. It is clear that Lean Start-up techniques have little role to play for these types of complex, critical systems.
Other robotic systems are not critical in the same way that military and surgical systems are, but they are critical all the same. Consider home care/lawn care products such as robotic vacuums and lawn mowers. These systems are typically purchased by technology early adopters, often at a high price. It is these same early adopters that push sales through word-of-mouth. For this class of robotics technology a minimally viable product will not serve.
Home care/lawn care robots cannot be made available for free (hardware is costly), and even if they could, poor performance would sink any chance of future market viability (a vacuum that does not clean, a mover that does not cut). The same holds for research and educational robotics.
It can be seen that criticality as described above is partially a reflection of the fundamental difference between software and robotics. Software typically operates in a purely virtual world. Robots are characterized by the sense-think-act operational paradigm with the ?act? occurring in the physical world.
Physical systems acting in the physical world incur a cost (hardware, complexity, increased development time etc.), as well as the possibility of causing physical harm to humans and property. Product adjustments cannot be made on the fly, thereby foiling the process of continuous deployment.
Don?t Forget the Hardware
Robotics technologies are composed of both hardware and software, and as such robotics start-ups are much more capital intensive than software firms. You simply cannot quickly build and make available to early adopters minimally viable robotics products. First, the costs would simply be too high. Even simple robotics systems such as smart toys incur substantial production and tooling costs even for small runs. Outsourcing production, which is the norm for consumer class products, requires additional costs (plus supplementary set-up charges for small runs and post contract changes).
Production, whether in-house or outsourced, also takes time, often a great deal of time. New manufacturing technologies and services could reduce costs and production times. For example, 3D printers could be used, but this would place limits on the types of systems that could be produced.
Smart Toys and Robotic Gadgets
So do Lean Start-up techniques have any applicability in the robotics space? To do so the systems must be very low cost, and incapable of doing harm to people, property and businesses. So much for robotic technologies that actually do useful work. In addition, the systems must be able to be produced rapidly and very cheaply. Again, serious systems need not apply.
It would seem that lean techniques can be applied only to a very select number of robotics systems types. Smart toys and other types of low cost consumer gadgets might fit the bill.
So too would be more complex systems that are based on proven, readily available, standard hardware. Such hardware could be leveraged using smart software and augmented with low cost materials that can be easily produced.
Dan Kara is analyst-at-large for Robotics Business Review. He can be reached at [email protected].