I have been looking into all the IoT Cloud Platforms available, AWS IoT, Windows Azure and Google Cloud Storage Solutions. Any suggestions on which one fits best for a start-up?
Any experience or feedback appreciated.
I personally prefer AWS which bundles a large number of different services you can use. Since they are from same vendor, you will get seamless integration between them. For example, in our use cases, we use S3, EBS, EC2, Cloudfront, Amazon Aurora, Dynamo, Elastic IP, and transcoding services. You will get not all but almost all services you need. Some Google services are good, e.g. Spanner we might use it later. Don't have much experiences of Azure.
It depends on your team and your budget . AWS is the most complete one but could be expensive compare to other solutions. In my opinion AWS is the leader of the cloud platforms however it has a higher learning curve (for your engineers) More over, estimating your price could be difficult. Some times for a startup with less resources you need other solutions less powerful but affordable : digitalOcean or linode
Does it have to be hosted on/in cloud? Costs and ease of migrating are important. I have most experience with AWS. Ask all of them for promotions.
1. They all have special "try it free for a while" deals for qualified startups. Really generous deals. At least they used to... and most likely continue to... with the right introductions. Talk to your local incubator people about what it would take to get the right introduction.
2. Architect your software in a manner that enables you to easily migrate from one platform to the other.
this will be long...
several considerations AWS vs Microsoft
1. Data transfer/communications of messages - both do so in different ways
Microsoft's gateway enables bi-directional communication between devices and the back end of the cloud -- meaning that data can travel to and from the cloud. for .e.g., Devices that are IP-capable, using Wi-F or Ethernet, will have two endpoints that interact with Microsoft's gateway to send messages to the cloud as telemetry data, requests for execution and the action of a command.
The Azure IoT Hub also enables endpoints in the cloud so that the back-end system can send messages to devices. These messages are timed so that they will expire in case a device receives a late command that is no longer applicable.
2. Pricing - Microsoft's pricing model revolves around to paying for the IoT Hub unit related to the number of devices and messages per day, AWS IoT's pricing model charges for the traffic of messages published from devices and delivered to devices.
for e.g. MS - S1 around $50~ / month up to 400k messages a day/unit; S2 $500~/month upto 6million messages/day/unit S3 - up to 300million messages/day/unit cost around $5k~ all these Ss editions plans enable a message meter of 4kb
AWS - bases its IoT pricing on a million messages exchanged policy, which means that the total cost depends on publishing messages to and from devices and applications. For instance, U.S. customers are charged $5 per 1 million messages published to AWS IoT. There is no minimum fee.
Both of them offers free trial..
Both offer open-source via github both different SDKs for diff languages and platforms.
Microsoft for e.g., .Net Framework UWP apps running on windows 10 etc etc. SDKs for Ansi c99
AWS for e.g., nodejs, arduino yun board, ANSI c99
AWS - HTTP, MQTT (official )
Microsoft - HTTP, AMQP 1.0(official)
- both ok but there are more details you can compare based on regulations you wanted to meet
- look for Microsoft Azure Certified.. lots of partner texas instruments, intel, raspberry, freescales ARM mbased etc etc
- AWS there are starterkit for you to rapidly prototype. AWS partners: qualcomm, intel, beaglebone.. boards for low-constrained devices like microchip
7. applications & services
both offer good array of services that come with the platforms ... collection and ingestion of data and beyond..
I own a start up company that is Microsoft approved Cloud solution provider | AWS web service partner network.
The answer depends on various factors like how much time and resources you want to spend. And how technical the team is.
AWS has a full offering on this and will be the easiest to implement considering the support and lot's of tech blogs done by a bunch of good technical folks. github will get you a lot of examples and starter material.
Google cloud on the other hand provides you the most robust and cost effective setup and a lot of daily free limits which should be good enough for your dev work but there's isn't that level of support or technical documentation freely available..
Microsoft Azure is the best for start-ups.
I found AWS IOT very easy to use
well everything has been already said here! so i'll just add that it all depends on your requirements however AWS is a bit technical and more suited for enterprise level organization Azure is more suited for SMBs. besides now is the best time to invest in Azure