Hardware, Software & the Internet of Things
June 23–25, 2015 • San Francisco, CA

Data rich and attention poor: Intention-based data architecture for intuitive IoT design

Abe Gong (Superconductive Health)
2:05pm–2:45pm Wednesday, 06/24/2015
Data
Location: Fleet Room (Bldg D)
Average rating: ***..
(3.90, 10 ratings)

Prerequisite Knowledge

No special expertise required.

Description

Typical IoT applications are data rich and attention poor: they must be able to follow the age-old rule of good design—“do what the user expects”—based on contextual data, instead of direct commands from users. Because of these constraints, IoT designs are powerful and intuitive only to the extent that their data architectures are capable of faithfully expressing user intent.

This session shares principles for building effective intention-aware data architectures. I draw on examples (good and bad) from across the industry, as well as personal experience at Jawbone and Metta.

Geared toward the practical needs of product development, this talk will address:

  • Articulating intention layers, and translating them into specifications for designers and engineers
  • Managing tradeoffs between expressiveness and simplicity
  • Building a culture and a common language around user intention, across different departments and disciplines
  • Finding opportunities for high-value data products
  • How intention layers work for devices with different levels of autonomy
  • Avoiding pitfalls in hardware and software architecture
  • Engineering around constraints in connectivity, battery life, and distributed computing.

More broadly, this talk brings together two active conversations in the IoT community—interaction design and data architecture—and argues that they should be the same conversation. The intention layer is the point of contact between the two approaches.

The presentation does not include a live demo, and no special expertise is required.

Photo of Abe Gong

Abe Gong

Superconductive Health

Abe Gong is fascinated by the intersection of data science, behavior change, and the Internet of Things. Prior to co-founding Metta, he was the first data scientist at Jawbone, and lead data scientist at Massive Health.