When Lao Tzu – the ancient Chinese philosopher, who seems to have spoken almost entirely in quotable quotes – said “Those who have knowledge don't predict. Those who predict don't have knowledge”, he obviously never foresaw the transformative impact artificial intelligence and machine learning would have. While prediction remains a far from perfect science, it is indeed closer to a science than it has ever been.
Take predictive analytics, for example. Predictive analytics finds patterns in data in a way that a mere human could never hope to achieve, using massive computational power and the incredible capabilities of artificial intelligence to provide previously-unachievable insights. The more data, the more likely those predictions are to be accurate – and if there’s one thing we’re not short of today, it’s data. Big data and predictive analytics go hand in hand. Data is turned into information – and information is turned into knowledge. And: with the proliferation of sensors – a key feature of the Internet of Things – we can collect almost all the data we’d ever need.
To grossly over-simplify: if a specific set of circumstances has led to exactly the same outcome nine times, the chances are that that set of circumstances will lead to an identical outcome the tenth time. Psychologists talk about how past behaviours are an invaluable indicator of likely future behaviours. Patterns emerge that are likely to be repeated. We learn from experience, because experience tells us what to do and/or what to avoid in the future.
All of which brings us to Abaco’s newly-announced Health Toolkit. It is, in effect, a tool to facilitate predicting the future. As with predictive analytics, it harnesses the data that is available throughout a typical computing subsystem or system; applies advanced software technologies to that data; and, based on its interpretation of that data, enables the prediction of potential system failure. For any mission, that is a vital advantage.
Part of Abaco’s growing NodeWare family of software/middleware tools, the Health Toolkit can help identify patterns of behavior that can lead to system malfunction. It ‘publishes’ system information on the network, and this information can be used in a number of ways. Firstly, it can be seen via a dashboard GUI or web browser such that a manual intervention can be made by an operator. Perhaps more compelling is the second option: the information can be consumed by the application software, enabling the application to take appropriate action in response. Thirdly, the information can also be collected and logged to a database for live or future analysis. All three of these are independent and can also occur in parallel if desired.
Raising the bar
The Health Toolkit raises the bar in terms of what’s possible in terms of preventing system failure. It does this firstly by taking a holistic approach to the system. Most individual embedded computing boards have their own internal diagnostics that allows them to be interrogated – but the Health Toolkit works, not in a piecemeal fashion at the component board level, but at the system level.
Secondly, the Health Toolkit is hardware-agnostic. It doesn’t just interrogate Abaco boards within a system: it recognizes that many of today’s systems are multi-vendor in nature – and is capable of interacting with non-Abaco boards too. That, we believe, makes it unique.
The Health Toolkit is also non-proprietary in that it’s built on the Data Distribution System (DDS) open standard middleware layer that’s used in many industries and aligns with wider common system architecture standards such as FACE and SOSA.
In summary: we believe that the new Health Toolkit has the potential to make a huge contribution to system reliability and mission success. Francis Bacon, an English philosopher from the 16th century, was the first to use the phrase: “Knowledge is power”. Today, despite anything Lao Tzu may have said, knowledge gives us the power to predict the future – and Abaco’s Health Toolkit delivers that knowledge.