Werner Vogels, CTO Amazon - Keynote
- Customer behavior in changing rapidly around speed of adoption or new technology products
- Today the amount of data generated in the 1st day of a child's life (ie multimedia) is equivalent to 70 times the content of US Library of Congress
- We are in Industry 4.0 already
- Example: Sushi chain in Japan using an automated assembly and client delivery line. Each plate of sushi is 'assembled' by machines and is uniquely identified through an RFID tag to track its freshness and automatically retract it when necessary
- Example of Analytics/IoT used in agriculture - The Climate Corporation (Total Weather Insurance)
- 'If you went to bed last night as an industrial company, you're going to wake up this morning as a software and analytics company.' - Jeffrey R. Immelt, GE CEO
- GE - the only company in the S&P 500 from the original ones due to continuously reinventing themselves
- Deconstruction - company specialised in adding sensors and analytics services to construction sites. Use cases include building an audit trail for governance and in case of future problems
- The Sports of the Internet of Things
- Every player has sensors in their training gear
- Team strategy today is not defined on the field anymore but in the analytics rooms of the coaches days ahead (ie tracking includes measuring if 'pulse' of the team is 'in sync')
- Socks for runners to track cadence, foot landing and foot contact
- Philips health suite: e-care platform built around health tracking, prevention and care
- Sonos- 'a 10 year product can do things that weren't invented 10'years ago' via software update capabilities
- AWS IoT:
- Scalability at core of its architecture
- Analytics: retrospective, here and now or predictions
- Amazon Kinesis used as live stream analytics engine (here and now)
- Use case - removing the 'tyranny of choice'. Refers to reducing the number of choices for your customers based on past data and other preferences to enable a better, straightforward customer experience. Ex: choose a toothbrush from 3-5 options instead of having to choose it from 52 options
Other takeaways:
- Evrythng - Dominique Guinard, CTO
- Started by 4 co-founders in 2011
- Pragmatically targeting a much larger market opportunity (80 Bn items and 5-10 Tn consumables compared to roughly under 30 Bn connected 'things') and 'inevitability' for simple, taggable consumer products to be tracked and interacted with.
- Doing load stress simulations: flood.io + gatling.io + EC2
- Amazon Alexa - Voice Interface / Personal Assistant: How to add new 'skills'
- An Alexa skill is a new piece of custom functionality that enriches Alexa's range of responses
- Can be created by anybody and published subject to an Amazon vetting process (similar to Apple Store)
- Alexa to be integrated by Ford and Kitchen furniture manufacturers in their products
- Alexa has a companion App which can provide the user with additional visual information
- Voice User Interface (VUI) best practices (see picture below)
- Crawl - Walk - Run Approach
- Make it as natural as possible
- Support multiple utterances
- Utilise the built in help intent
- It takes about 3 weeks on average to build production ready Alexa skills
- Tip: spend more than 50% on designing the interactions
- In my opinion Alexa is an unavoidable step forward in how we interact with technology but it is in it's infancy at this stage.
It is still quite limited - i.e. no support for recognising different voices and associated profiles, no monetisation model yet to name but a few. And it looks more like a good speech to text and text to speech service rather than anything else at this point in time.
I would expect this to change however – especially due to the straightforward and fast process of building new skills for Alexa.
No comments:
Post a Comment