While the IoT sector’s never been short of visions, turning those dreams into a working – and profitable – reality can still be problematic. When it comes to choosing an appropriate IoT strategy and the right partner to support them, companies are faced with a potentially bewildering array of different vendor solutions – all currently described under the convenient, if maddeningly imprecise, generic term of a ‘platform’.
While clarity is slowly starting to emerge in terms of a more standardised lexicon of functionalities and features, many of the proposed platform solutions currently out there carry with them the individual baggage of each of their vendor owners’ pasts. In practice, this translates into platform development strategies that exploit the vendor’s historic investments in technology and try to adapt them to the needs of present day customers. The end result is all too often a system that’s not truly designed for the environment that it’s going to end up being used in.
It’s against this backdrop that M2M Now’s editor, Alun Lewis, recently sat down with Dave Friedman, co-founder and CEO of relatively new on the scene company Ayla Networks to talk about his company’s approach to resolving the ‘platform’ conundrum. The focus for Ayla is on helping their manufacturer customers make the right choice of platform to enable rapid product and service innovation – not limit it.
M2M Now: Dave, in our initial conversation, you flagged up how inappropriate and imprecise terms and definitions were hampering our sector’s progress and the wider adoption of IoT strategies. Can you expand on this a little?
Dave Friedman (DF): We all have a natural tendency to draw on the past to explain and interpret the present – and predict the future. This tendency also applies equally to what you might call the institutional memory of companies. As the IoT landscape has expanded rapidly over just the last few years, creating a huge variety of different vendors, old and new, any analysis of their solutions soon shows that they reflect the heritage of both the companies concerned and the executives running those companies.
In the context of the IoT, this presents some serious problems. For a start and at the most basic level, one primary challenge involves our industry’s use of the word ‘Things’. There’s the dictionary definition of the word – but then your individual definition of ‘thing’ in an IoT setting is going to depend on what industry sector you’re in, what product or service strategy you’re trying to drive for your company, what relationships you want with your own customers and users and so on.
Similar issues apply to the word ‘platform’. For some vendors this automatically implies that there’s some form of controlling or filtering physical gateway sitting out in the network. For others, it implies that there’s going to be one single, all-encompassing operating system, tying everything together into some seamless utopian vision – and we all know how successful Microsoft has been using this strategy in the mobile sector.
Now, myself and my co-founders are just as guilty as anyone else of drawing on previous experience –though in our case I’d argue that that’s turned out to be serious advantage – especially if you look at our current customer base. We are fortunate to work with the leading manufacturers of everything from home appliances, fans, and fire and safety products to water heaters, water softeners, boilers, and heating, ventilating, and air conditioning (HVAC) systems.
M2M Now: So what insights has this previous experience given you?
DF: I think one vital differentiator here is that I and a number of my close colleagues started literally from the ground up – which in our industry starts with the silicon and actual chips that make up the building blocks of today’s connected civilisation, rather than from imposing some top-down systems perspective. That’s meant that we’ve seen no need to reinvent or reuse the middleware sort of concepts used by many ‘platform’ focused players and instead looked to link chips directly with the cloud in the most elegant and efficient ways possible. In most cases in the IoT world, you do not have to deal with huge amounts of data coming in from each individual device and these days we have cloud-based architectures that are implicitly able to handle scalability issues in secure and cost-efficient ways that weren’t there even just a few years ago.
By putting Ayla-enabled ‘hooks’ into a wide range of chips, which then in turn can be easily integrated with an almost infinite range of systems and devices, we’re able to put control of their own destiny back into the hands of product development engineers. For example, in the first days of the company, we met with many leading Wi-Fi chip set manufacturers – Broadcom, Qualcomm, Marvell and others – who quickly saw what we were trying to do and wanted to work with us. Since then, we’re continuing to broaden our range of connectivity options and now also have added Bluetooth to our portfolio. One early endorsement of our strategy came from Cisco who quickly understood what we trying to achieve and they became an early investor in us.
M2M Now: Insights like those might be theoretically elegant in engineering terms, but how do they translate into actual practice for your own customers and who are your target markets?
DF: At the risk of adding another key differentiator to our list, I’d like to suggest once again that it was our heritage in being closely involved with innovative engineering principles across a number of different sectors that’s also aided us. While there’s a huge amount of hot air expended about strategies to handle disruption and innovation, we clearly saw our role as helping large manufacturers get connected without having to reinvent the wheel each time themselves – so we’ve typically been targeting Fortune 100 type companies – with significant success, as I said earlier.
Every major manufacturer on this planet now recognises the need for an IoT strategy – the question now lies in how best to implement it. Everyone who’s ever worked for or with a big corporation knows what a minefield any big shift in strategy creates. Budgets, headcounts and even executive parking spaces all get fought for during the associated disruption and political uncertainty and the end results are often commercially and technologically underwhelming.
A number of our big manufacturing customers did start off by trying to set up IoT teams and departments themselves, but soon found that the skills required were in short supply and the investments and long pay-off periods involved would quickly impact their core focus and their own product R&D. Indeed, in a few cases we did actually go through a ‘bake-off’ with internally developed products – and I’m pleased to say that Ayla won against the incumbent teams, to the longer term good of the customer.
After putting a toe in the water, so to speak, before rapidly pulling it out, manufacturers realise that our approach seemed to make obvious sense as an alternative strategy, especially since we’re able to draw on our existing close relationships with the chip manufacturers and deep understanding of their domain. Using our Ayla Embedded Agents, product development teams have access to a fully optimised network stack with the additional protocols needed to connect their devices to Ayla Cloud Services. Time and cost to market gets cut and, with the ability to ‘hook’ any device, sensor or control interface into the Cloud environment, a customer’s engineers can get on with adding value to their own product portfolio.
M2M Now: So what’s your strategy for application development and integration?
DF: That control side is also obviously important, so complementing our agents and cloud are our Application Libraries. Just as our agents simplify the integration of connectivity with devices, our Application Libraries do exactly the same for those interfaces where the IoT meets its human controllers – via smartphones, tablets or the Web, through both Android and iOS devices.
We’ve also taken the control aspect one step further, with our Agile Mobile Application Platform (AMAP). Built upon Ayla’s mobile software libraries, AMAP provides pre-made, pre-tested software code that supports the main features that users expect from mobile app control of a connected device. That includes things like sign-in, registration, device setup and control, password recovery, wireless setup, schedule creation and management, support for push notification and timer setup. It also includes robust security features. Manufacturers of connected products can license AMAP and have around 85% of their mobile app coding and development work already done. All they need to do is specify the 15% of functionality that personalises and differentiates their products and allows them to “skin” the app with their company branding.
AMAP exemplifies Ayla’s fundamental understanding that, sure, a manufacturer could hire teams of engineers and buy tons of equipment and do their own cloud connectivity -and some of the bigger manufacturers do just that. But, for the vast majority of manufacturers of the variety of ‘things’ connecting to the IoT, it makes much more sense — from a financial, time and quality standpoint — to focus on the core capabilities of their products and leave the connectivity details to experts like us.
M2M Now: The term ‘cloud’ is another very loosely used word that covers a multiplicity of different flavours and approaches. What role is it playing for you and what’s your take on its evolution?
DF: There’s certainly a tendency to resolve every IT issue by saying, “We’ll do that in the cloud” and while that’s true to a certain extent, certain caveats apply – and we’ve taken those caveats into account in developing our own solution.
Most importantly, while the cloud idea sounds nicely global and is often explained that way, it is in fact intensely regional and it looks like it’s going to stay that way for a long time yet for some very good political, social and economic reasons – at least by those who decide these things.
When it comes to supporting manufacturers, wherever their HQ is based, Ayla has a probably unique advantage in having our cloud environment operating in multiple regions including China, where we’ve already got over 30 people working. If you’re a Chinese manufacturer looking to enable connectivity to the rest of the world or a Western manufacturer looking to simplify R&D with a partner of manufacturing resource in that country, we can provide a seamless domain.
In Europe, which again for historic geopolitical reasons has a very strict regulatory regime concerning cross-border data transfers, we’re about to launch a dedicated cloud service there this autumn. Once again this simplifies R&D and product testing and verification and enables our customers to build B2C relationships in cost-effective and secure ways across an entire continent.
Agility, time to market, risk reduction and simplifying integration and product management from a global perspective is just some of the key criteria for any platform that’s going to be fit for purpose in this second decade of the 21st century. We’d like to think that we fulfill all these needs – and enhance them further through our close links with key industry players like Cisco.
David Friedman is the CEO and co-founder of Ayla Networks. Prior to founding Ayla Networks, David served as VP, business development for ZeroG Wireless, a company providing low-power Wi-Fi products to the embedded space. Prior to ZeroG, David was an early employee at Matrix Semiconductor, serving in various sales and marketing roles before Matrix was acquired by SanDisk. In earlier roles, David helped drive pricing strategy at Intel, and was an analyst in the mergers and acquisitions group at Chase Bank. David has an MBA from the University of Michigan and a BA from Colgate University; he holds five US Patents.