In this session, Distinguished Engineer, James Gosling, discusses how AWS innovates in the Internet of Things. James shares stories and experiences in deploying IoT systems, and how AWS thinks of scalability in IoT. In addition, James shares his experiences in engineering Java embedded systems in IoT.
Report
Share
Report
Share
1 of 30
More Related Content
Engineering the IoT at AWS - IOT402 - re:Invent 2017
10. Real time ≠ real fast
•Predictable, Dependable
•No latency spikes
•e.g., HashMap considered harmful
•Throughput can be compromised
•All is negotiable
10
15. COURTESY OF LIQUID ROBOTICSCOURTESY OF ROGER DANNENBERG COURTESY OF MARC DONNER
Weird
McBlare, the robotic
Bagpipes:
Timing is critical
15
16. COURTESY OF LIQUID ROBOTICSCOURTESY OF ROGER DANNENBERG COURTESY OF MARC DONNER
Ivan Sutherland’s
Mechanical Walker:
It’s a giant
mechanical
cockroach
Weird
16
17. Weird
COURTESY OF LIQUID ROBOTICSCOURTESY OF ROGER DANNENBERG COURTESY OF MARC DONNER
Liquid Robotics Waveglider
Processors everywhere
(3 in the rudder module alone)
17
18. What we think about at AWS
• Scale
• Security
• Reliability
• Evolution
18
20. Scale: MQTT
• A common question “Why should I use MQTT? I’d rather
just open a socket and write!”
• Common encapsulation and routing framework
• Allows common messaging infrastructure
• DIY payload format
• AWS deals with secure scaling, so you can concentrate on
your device
20
21. Security
• A major pain point in early
IoT systems
• A major motivator
for AWS
• Cloud skills carry over
TLS, Certs, Auth, Crypto…
21
24. Reliability
Achieving reliability in a device is different
than in the cloud
• You can’t fail-over to another processor
• You can’t expect a god-like human to rescue it
• It’s in amongst regular humans, who do the
strangest things
• Code coverage is never good enough
24