VCs have woken up and smelled the robo-coffee. A bottom-up estimate for the US economy alone shows close to $1bn in VC investments for 2015, compared to approximately $200m in 2011. That excludes non-US deals and other borderline areas such as IoT, AI, 3D printing – so whatever definition of robotics we choose, we’re talking $1bn+ in VC investment. Yet clearly this is still a relatively ‘young’ space, in that there are as yet not many mega-exits, and the amounts invested still pale in comparison to the digital tech/mobile space. It is not yet clear to investors what may be the winning ‘formula’ that would enable a scale-up of investment flows in the robotics space. Nevertheless, should a mega-exit in this space materialise – such as the ongoing $5bn takeover bid for Germany’s Kuka Robotics by China’s MIDEA Group – the investment case for robotics companies will crystalize.
Economic history is full of examples of how the emergence of a new technological paradigm leads to evolution or disruption of winning business models. History also shows that such changes can happen quite fast if fuelled by capital (think electrification, rail, automotive, semi-conductors, E-commerce, sharing economy, and so on). So if we see robotics as a technological paradigm that is in the early stages of diffusion, then we can expect similar enabling elements to emerge around this technological paradigm, and shifts in the way of existing patterns of doing business, or even requires institutional change. One key ingredient is an investment model appropriate to the technology opportunity to effectively monetise such opportunities. Its nicer to call this a ‘recipe book’ than to use boring economist terminlogy, so I will stick with that.
(But if you’re inclined that way, here’s a great tome to get you started on your evolutionary economics and systems of innovation studies)
Previous booms in VC investments have been to a large extent facilitated through the emergence and acceptance of such recipe books about how to execute investment and growth in a particular space, the levels of capital committed, perculating up to VC fund size and average investment tickets. Clearly the investment models differ between semi-conductors (1980s), biotech (1990s/2000s), E-Commerce and Internet (2000s), mobile and sharing economy (2010s). Each of these waves saw the new technology paradigm interact with and build on – or disrupt – the pre-existing ways of interaction with existing standards, hardware and software infrastructure. At the start of each of these paradigmatic changes, there is a huge variety of market applications where the new entrants can apply their recipe. This makes the potential success of investments exciting, as a successful investment ‘recipe book’ can be applied across a wide spectrum of applications.
Emerging Recipe Ingredients
To stick with the metaphor, its clear that many different recipes are being tested right now – some more bitter than others. However, the rapid growth of VC investments makes it more likely that a set of winning investment recipes will emerge in the next 2-3 years, as failures and exits multiple. So its worthwhile to take a look at current experiences by the new wave of VC investment in robotics, as investors identify some emerging traits. Below is a list of some candidate recipe ingredients:
- Modularisation: the increasing modularisation of key robotics build components, such as open source operating systems, hardware component CAD libraries, off-the-shelf components, ever cheaper and more powerful sensors, each and in combination provide robotics startups with a growing bank of solutions that can speed up build time and iteration.
- Lean startup: In turn the faster and cheaper build and iteration possibilities make it easier to apply Lean Startup principles by many robotics startup. The use of robotics outside of the factory really opens up the use of Lean Startup methods, such as rapid design iterations and A/B testing, to rapidly develop the design of the product systems and commercial models that go with it – and cut time-to-market
- Connectivity and IoT: The rapid growth of the Internet of Things provides robotic products with a connectivity infrastructure to plug into.
- Hardware as a Service: The connectivity in turn enables scalability in the deployment of systems similar in some ways to SaaS, as units can be upgraded remotely, and also monitored for levels of usage and breakage.
- Unmanned Vehicle Platforms: The vast number of unmanned vehicle platform types, whether ground, air or sea, are multiplying the application areas where a robotics solution may be feasible, by opening the scope for a move from unmanned remotely controlled, to semi-autonomous, and eventually autonomous
- Artificial Intelligence: Major advances in computing power and cumulative advance in AI fields including machine vision and deep learning are bringing down the cost – and speed – of development of new products
Example: Q-Bot – your friendly Victorian house underfloor insulating robot
I will illustrate some of these factors with reference to one of our portfolio co
mpanies. Robotic solutions are uniquely placed to enter and scale into multiple markets because of their ability to manage jobs that are too dangerous, complex or boring for human operators. London-based Q-Bot is a great example of a robotics company that has built a robotic solution around precisely such an application problem. Q-Bot develops intelligent robotic tools for the building industry that turn difficult, disruptive, tedious and dirty jobs into clean, efficient and safe processes. The company’s first commercial application is designed to perform underfloor insulation for aging Victorian properties (of which in the UK we have many!). See great videos of the product in action here
The recipe ingredients I listed above are all present, and I think key to the company’s growth. From the outset the team invested in rapid production in-house facilities including CNC machines and a high-end 3D printer. In combination with off-the-shelf products, this enables the company to rapidly build and iterate the design of a family of robots in response to real experience in the field. They have integrated a range of high-end sensors to provide an extremely intelligent solution for a very low-end area of building – enabling them to redesign the service package a client gets. Continuously improving machine vision computing and control systems provide the company with economics of scale and scope, as the system becomes ever more sophisticated on the back of real life experience.
Where are the big investment returns?
I see several opportunities for superior investment returns in the robotics space. Broadly speaking these are in end-user B2C products, components and solutions providers to the robotics value chain, and B2B robotics solution providers. In the B2C space, followers of the industry are keen to find the robotics equivalent to the iPhone – that elusive package of cutting edge technology, design and ‘fit’ with consumer needs that can accelerate adoption. As the industry growth accelerates, and market uses increase, businesses who are able to place themselves as the providers of key components such as operating systems, artificial intelligence software modules, power drives and Internet of Things modules that help the industry scale will also provide significant investment return opportunities.
At EcoMachines Ventures we see the most significant B2B opportunities in business applications that can service, or assist certain worker tasks or provide infrastructure support, via remote maintenance, monitoring or fixing components. We look for startups ‘hyperfocused’ on a specific market niche, where the integration of best in breed technology, and deep knowledge of a specific, real application with a high-payout can provide the basis of an exciting, high-growth business. Often the application may be in a relatively ‘boring’ business area, but this is where some of the highest potential for productivity gains – and investment returns – lies. We also look for capital-light models that make use of what’s already out there, and also for business plans that obsessively search for early cash flows as a way of validating market demand.
…And one more thing
As I write this sturdy blog on a Sunday, my wife observed en passant with (I think) a degree of sarcasm that a DIY robot would come in handy in disintermediating me from the household. Fine by me, as long as an AI bot doesn’t disintermediate me from writing witty blogs!