- Collection of three coordinated, core technology layers to make connected devices
- Thread: IoT network protocol Google helped get started (and whose organizing group has many companies, not just Google)—creating a standard for mesh-based IoT connections
- Weave: API layer for IoT with a micro-OS (here’s a bit more about the implications, from Verge)
- Brillo: IoT-optimized variant of Android for embedding in devices (again, a bit more from Verge)
Smart/Intuitive Technologies (Predictive)
- Products and services with a data layer which makes the experience feel intuitive to the customer—a minimum of effort and thinking is required to get value
- Usually requires integration of many systems and datasets
Examples:
- Recommendation Engines
- Google Now
- Predictive traffic calendaring (Waze travel timing)
SDKs & HDKs
- SDKs enable developers to quickly develop fast, stable, secure and consistent software for end users
- Large companies provide SDKs to smaller companies developing on their platform for more consistent user experience
- Takes form of toolkit of software elements (sometimes as a library or framework) and documentation; may have multiple contributors (in open source approaches)
Examples:
- HomeKit, HealthKit
- Apple Human Interface Guidelines
- jSON
- Webkit
Predictive Analytics
- Use of analytics tools to predict customer or market behavior, other factors
- Requires extensive algorithm development
- Can incorporate developer biases if not properly constructed
- Best when integrated with active feedback loops from customers
Examples:
- The back end of recommendation and suggestion engines
- Tableau
- SAS
Open Source
- Software developed by multiple parties
- Shared/non-ownership of code
- Promotes innovation by giving an incentive for developers who benefit from entire ecosystem
- May be difficult to keep consistent and may have hidden technical costs
- Can rapidly accelerate development across coopetitors (cooperative competitors) a la Linux
Examples:
- Linux
- Thread Group
- Darwin
- Raspberry Pi
- OpenStack
Internet of Things
- Connection layer for physical devices, such as industrial or home sensors, thermostats, etc
- Interoperability, security and reliability of IoT systems may be a barrier to entry
- Will evolve into a Social Network of Things (SNT) where devices can better connect to and negotiate with other devices; see causeit.org/SNT for more information
Examples:
- “Smart” power meters
- Phillips Hue
- Nest Thermostat
- Withings
Integration & Interoperability
- Key element of connected products
- Use of Application Program Interfaces, Software and Hardware Development Kits, shared standards, libraries and frameworks
- Requires ability to selectively share and potentially revoke data access
- Requires clarity on where decisions are made in a chain of connected services/devices
Examples:
- USB
- APIs
- SDKs & HDKs
- Apple HomeKit
- Facebook API
Data Rivers & Data Lakes
- Data Rivers: pipelines of data
- Data Lakes: aggregations of multiple data sources, either with or without data validation
Examples:
- Real-time analytics performed on mass data sets
- Facebook Social Graphs (aggregate)
Data in Motion
- Data which changes often, such as a live feed of video or an algorithm-driven analysis of a stock market
- Dynamic
- Secured multiple ways, such as securing the ‘pipes’ the data flows through and through authentication of users authorized to access
- Data architectures built to be synchronized across systems (central data or self-reconciling systems)
Examples:
- Google Docs
- Video feeds
- Algorithm-based data sources
- Cloud-based systems
- Real-time analytics
Data at Rest
- Data which, like a file in a file cabinet or row in a spreadsheet, stays the same until a user modifies it
- Static
- Secured usually by securing the perimeter around it (locking the file cabinet, password-protecting a spreadsheet)
- Hard to synchronize across systems (copies and versions)
Examples:
- Spreadsheets
- Hard copies
- Attachments on email
Big Data
- Collection and aggregation of large pools of data for analysis
- Many companies pool large quantities of data without knowing use cases yet
- Storage, privacy, security, speed and reliability are major concerns
- Big data + analytics needed for business insights; multi-sided platforms needed for advanced data-driven products
Examples:
- Collection and aggregation of large pools of data for analysis
- Many companies pool large quantities of data without knowing use cases yet
- Storage, privacy, security, speed and reliability are major concerns
- Big data + analytics needed for business insights; multi-sided platforms needed for advanced data-driven products
Artificial Ingelligence
- Augmentation or emulation of human intelligence in machine systems
- Creation of new, non-human forms of intelligence
- Includes expert systems (modeling human knowledge) as well as systems with decision-making skills
Examples:
- IBM Watson
- Assistant.ai
- Autonomous driving systems
APIs
- Application Program Interfaces to create conduits between various systems and datasets
- Enables internal or third-party developers
- Requires robust and secure architecture, including authentication, data integrity efforts and use logging
- Necessary for platform-based value
Examples:
- HomeKit
- Facebook Connect and other Single-Sign-On (SSO) options
- Ripple
- Fidor
- Apple CarPlay
- Software Development and Hardware Development Kits (SDK and HDKs)
Analytics
- Discovery, interpretation and communication of meaningful information in data
- Multidisciplinary—covers entire methodology and data value chain
- Many subfields, such as text analytics
- Temporal (time) element: batch (past) analytics, real-time analytics, predictive/prescriptive analytics
Examples:
- Google Analytics (web properties)
- Sales analytics
- Inventory analytics
- Mixpanel