MachNation is one of the world’s most respected IoT analysts, as they actually evaluate IoT system performance by hands-on testing, to guide companies who are considering adopting new IoT platform(s) and components. Most MachNation reports are available for a fee but they have recently released for free their 2018 analysis of one of the biggest IoT platforms: AWS-IoT.
Each MachNation platform report is based on more than 100 tests, with 3 scores for each:
- Completeness (did it do the job?)
- Sophistication (good docs, well-designed UI?)
- Ease/Timing (time to achieve the task).
Here we summarise the parts of MachNation’s AWS IoT analysis which are relevant to you if you’re considering adopting DevicePilot as your analytics solution. DevicePilot is an IoT analytics solution which complements IoT platforms like AWS-IoT by providing much easier-to-use Analytics, Monitoring, Visualisation and Automation, delivering much quicker time-to-insight for any modern business. DevicePilot is a certified AWS partner which integrates with AWS IoT in seconds. It requires no coding, IT or data analysts for integration or daily use and is ready-to-use software that every business person can use for visibility, monitoring, operations and business automation.
Since this summary is necessarily selective, we strongly recommend reading MachNation’s full report for a complete picture - see link at end.
There now follows our summary of MachNation’s AWS IoT report in each relevant category, with summary and highlighting added by us.
AWS is “flexible and extensible” but “still somewhat lacking in overall usability”. “QuickSight help[s] improve usability but offer fewer options for extensibility and overall capabilities”
Test C-04-01 (“Configure rule/alert associated with device status”) took 15 minutes to achieve this goal, with only a “1 out of 3, somewhat unsophisticated” experience. In contrast this task can easily be achieved in DevicePilot in <1 minute, in most cases.
“...AWS IoT...somewhat deficient in terms of data-management operator usability...lack pre-configured operator interfaces or dashboards, instead relying on … competency with SQL-style syntax or management through third-party tools. This can make casual visibility into stored data a less than smooth experience…”
Test C-02-04 (“Compute aggregate statistics for multiple data points”) and C-02-07 (“View historical sensor data for a group of devices”) each took 20 minutes to complete - for tasks that take only seconds in DevicePilot.
Event processing is generally considered to be the real-time processing of streaming data. MachNation considers that AWS IoT has “multiple effective avenues for achieving complex event processing...Rules service provides a flexible option for routing data...using a SQL-like syntax, allowing operators/administrators to filter incoming data streams in real-time, and either trigger actions or route data based on matching conditions.”
Test C-04-01 (“Configure rule/alert associated with device status”) and C-04-05 (“Configure data action based on data configured range exceeded”) each took 15 minutes to complete a task, albeit one which can be completed in seconds in DevicePilot.
This MachNation category seems principally to be about how easy it is to get data out after analysis, and to integrate “sideways” with other business tools (i.e. everything except ingestion). This is an area where cloud providers often don’t generally score very well, at least out-of-the-box, as they really want you to stay and do everything within their Cloud.
MachNation considers that “External integration is an area of weakness for Amazon … lacks integration with off-platform services… Salesforce… is the only integration provided.”. In contrast, DevicePilot provides several easy ways to integrate with external applications such as BI and CRM tools.
MachNation says that “Monitoring of IoT assets and data streams is an area of mixed strengths for AWS IoT” and that “...interfaces are designed assuming a limited number of devices being managed through the UI”. “Visibility into device state/status is also somewhat limited”, “...limited bulk device views...”, “...omission of any user-editable dashboards or interfaces...” , “...customers will undoubtedly need to develop and build their own operational interfaces for all but the most simplistic IoT deployments.”.
The problem with “building your own operational interface” is that this wires-in engineering cost and delay for ever-more. Whereas DevicePilot provides a much more cost-effective alternative operational interface - instantly.
Tests C-07-06 (“View historical device status of a single device”) and C-07-08 (“View device status for multiple devices”) each took 10 minutes to achieve, whereas DevicePilot makes these vital everyday tasks a snap.
“...Amazon excels at developer usability” but “The most deficient aspect of the Amazon platform is the lack of pre-built operator interfaces, instead of requiring customers to develop and provide these interfaces” - which highlights the gap that DevicePilot fills very well.
Test C-08-07 (“Evaluate overall cogency/quality of operator user interface (UI)”) achieved a low score of 1 (“partially complete”).
The MachNation report paints a picture of AWS IoT familiar to anyone who has used AWS’ other Cloud services: they’re very good at infrastructure but don’t really “do” applications. AWS IoT is a technically very competent and secure service, but one which is not very easy or quick to use, even for highly technical people, perhaps best summarised by the following sentence from MachNation:
“AWS IoT is an excellent solution for customers looking for a data-ingestion and aggregation endpoint for their IoT solutions, yet still lacks some of the comprehensive features required for more complex IoT deployments.”
This matches our own view very well: For a secure, reliable, scalable, good-value IoT cloud broker, we highly (and frequently) recommend AWS IoT. DevicePilot then provides a highly complementary analytics application to deliver analytics agility in your organisation.