Sunday, April 30, 2017

Alfidi Capital Attends Bluetooth World 2017

I have always been curious about this Bluetooth stuff, so I had to check out Bluetooth World 2017 down in Silicon Valley. I did not actually see anyone with blue-colored teeth at the conference but maybe I just didn't look closely enough into people's mouths. There's more to Bluetooth tech than just drinking blueberry juice.

Alfidi Capital got a free pass to Bluetooth World 2017.

Some Cisco guy supposedly predicted that the IoT market would be worth over $14.4T by 2023. It's one of those phantom quotes that gets thrown around at these types of conferences. The Bluetooth Low Energy (LE) aficionados here were all about grabbing that big IoT market. There's no free lunch with this tech, meaning there's a tradeoff between longer range and higher bandwidth forcing developers to choose between optimizing for either speed or power.

The "mesh networking" protocol enthusiasts here think it's some revolutionary leap for power optimization over many IoT devices without recourse to gateways or routers. We hear about revolutionary leaps all the time in this valley. IMHO developers for Bluetooth and others techs are way too enamored with smartphone I.D. authentication. Techies never want to admit that bad people can steal or hack phones. Once again, everyone's pitching convenience over security! The immaturity in this valley is mind-boggling.

Bluetooth dominates the wireless audio market and it will probably rule the beacon market in retail store promotion. Online shopping is decimating retail, so beacons have a closing window of opportunity. The next recession will see this window slam shut, and I predict only a few surviving upscale retailers will truly leverage store beacons.

The big push for Bluetooth Mesh as an IoT architecture is doomed to fail if it relies on end users to implement security. It is also unclear how Bluetooth IoT devices and apps will generate revenue. Bluetooth SIG whitepapers should spell out the Cloudonomics metrics that will guide investment. Beacon data generating contextual information about customers will have commercial value once it is aggregated in clouds.

Let me continue on this money metrics theme. Energy use and data processing throughput are measurable costs for IoT devices. Any value Bluetooth adds must exceed the energy and data costs per device. It's also worth asking whether Bluetooth promoters intend the tech to be a completely open source system like ARM, Arduino, and Raspberry Pi. The choice to keep it moving in that direction will affect how developers build apps.

Bluetooth IoT devices will generate enormous volumes of data and metadata, requiring a greater role for machine learning in analytics. Developers should define where machine learning belongs in the data flow from device to cloud. I suspect only a minuscule amount of AI can reside at gateways due to power and processing limits, especially as Bluetooth Mesh pushes these support functions beyond gateways. Developers should focus their AI efforts on the cloud, where power and processing are theoretically unlimited.

It is unbelievable that some people at Bluetooth world think that blockchain will add value to Bluetooth in IoT. The vast amounts of data flows I mentioned above will quickly overwhelm a blockchain ledger. Imagine the constant forking and multiple layers of ledgers to track the new forks. Just say yes to machine learning and no to blockchain.

I foresee a market opportunity for startups based on scraping, collating, and cleansing data from Bluetooth Mesh networks. Venture investors should ask how these startups translate data from Bluetooth and other tech systems into a single format, as long as Bluetooth has competition. Startup IoT devices are amenable to crowdfunding if they have some consumer hook, like a cute beacon base station that invites in-store selfies. I am certain we will hear a lot more about AI and IoT convergence as startups figure out the combo is a hook to get VC funding.

I keep seeing startups jump on the wearable tech bandwagon and Bluetooth World 2017 had its share of starry-eyed wearable devotees. There is very little chance the wearable train will ever leave the station, folks. Wearables will always be way too expensive for most consumers. Battery life is a major limiting factor. Embedded sensors and processors make the products so fragile that they will need delicate care and frequent replacement. The connectivity path loss by distance from embedded devices to sensors is a challenge for devices hanging on a human body. These things will never be a mass-market product, and any personal data aggregated from wearables would likely pose HIPAA violations if released without a user's permission. Security matters "over the air" for wearables.

I asked one tech expert here about how we can design physical plant for maximum flexibility with future digital architectures. The guy said bandwidth is the bottom line, with physical structures built for the maximum possible access points and transmission capacity. There you go, Bluetooth fans, just develop your devices to fit in every building's joints, corners, and shafts.

Anyone who wants a head start on the next big thing should start adding Bluetooth knowledge to their skill set. The IoT frontier is like a gold rush and vacant land rush combined. Proliferating Bluetooth tech that identifies devices should make QR codes obsolete. Startups building solutions that can flex between Bluetooth and other techs will own the IoT future. You heard it here first from Alfidi Capital, thanks to Bluetooth World 2017.