Episode 4: Internet of Things (IoT)
In this episode Andrew and I talk about my favorite topic, the Internet of things (IoT).
Thinking about an IoT project or working on one? Listen to this episode first.
We discuss the most awesome combination of technologies AI, blockchain, and IoT. We also get into why open source is critical to IoT.
As well how cryptocurrencies fit into all of this. Build thoughtful IoT. - Fahad
Internet of Things (IoT)
IoT is transformative. Capturing and leveraging the extraordinary amount of data being produced by devices now is power.
Cleveland Clinic Foundation - beds are reporting triage patient stats back to nursing stations to stop things like heart attacks earlier.
Cars are reporting road and weather conditions.
Sensors will be (already are) massively important to our daily lives.
4:23- Hardware and software considerations in IoT
Wifi, bluetooth, 5G etc are solved problems.
Getting some of this tech to the last mile is important.
Cost to collect data from 100/1000s of devices needs to decrease.
Increasing computing power on-device is important for the future.
Basic misunderstandings of the science in the IoT space...standards battles are being fought everywhere because big players haven't looked at the work already done; they're just redoing it, many times.
Software has unique failure conditions but they're known. Hardware can fail in somany more ways. Hardware is the limiter, more and more.
Maturity in the IoT Market
8:20- What and who is talking in the market?
Explosive growth coming up.
Open source is critical to the success and growth of IoT
Open protocols are being adopted faster and wider (vs. closed protocols like Apple's HomeKit)
Speed of innovation is a huge benefit to robust open source communities.
Kubernetes vs. docker etc, for example.
Outta Control Data Firehoses
13:02- Data proliferation is an exponential curve; how do we handle it?
Build more datacenters?
Rockwell Automation, for example, collects petabytes of data a day.
Solutions include streaming data (sorting data)
A fewer private cloud implementations.
This level of data collection, though, is the key to unlock the power of machine learning or predictive functionality.
Applications of AI
17:00- They're everywhere; just need a bit of imagination.
Nest and other home uses.
Shipping (blockchain and AI)
Banking and ATMs (Diebold example)
Important to understand the tech as much as possible.
Skiplist has a world class understanding of the software domain of IoT. This frees customers up to focus more time and energy on their problem domains without worrying about the software as well.
Security and IoT
21:04- How do you address security and privacy in IoT pipelines and applications?
Security isn't done well right now. Many devices communicate unencrypted right now.
Privacy for users starts with howyou collect data.
Apple is a good case study for privacy and security in IoT with Siri-equipped devices.
Business incentives matter especially in this area. Ad-supported businesses aren't going to value privacy equally with other orgs.
Intelligent edge- pushing compute on data further out, away from servers and towards edge devices.
IoT and New Business Models
24:56- The money floodgates are opening
New networks and layers
Primitives are being constructed right now to begin expanding into new products in existing companies.
RFID and supply chain tracking.
To get started: bring domain experts to the table to communicate. Multi-disciplinary teams need to stay in lock-step.
Considerations Around IoT
28:53- What should we be watching out for?
The hype. Too many efforts to force it into existence. The tech isn't there many times.
Thinking that your problem can't be solved - start thinking about potential adoption of IoT or IoT projects early and take steps. Get people in the room to hash out the whats and hows.