1.      What’s the big deal?

Data has traditionally been collected and saved in databases, often relational databases, which have the capability to store large amounts of data. However, these databases have limitations due to the complex nature of data and its connections in the real world.

To overcome these limitations knowledge graphs are used. Knowledge graphs offer a novel approach to data storage whilst accounting forthe complex relationships in data. This results in easily accessible data, where it is possible to uncover hidden features and find new insights from your data.

‍

2.      What is a knowledge graph?

Knowledge graphs are models of data about a certain topic. These topics can be anything where data can be collected, such as people across multiple organisations, products for sale in a business or movies, actors,directors and how they are all connected. These models allow us to visualise the way connections are made when the data is used in the real world.

A knowledge graph composes of nodes, edges and properties. Edges and nodes are crucial to a knowledge graph whilst properties provide additional information.

·       Nodes are usually entities, such as people, organisations or products.

·       Edges are the relationships between nodes. Relationships could be between two nodes describing people such as ‘related to’ or ‘employed by’.

·       Properties can be any further information about a node and properties can vary depending on the node type. Properties donot link to the edges.

When we combine nodes, edges and properties we have a knowledge graph!

‍

3.       Movie Knowledge Graph Example

 A simple knowledge graph example is a movie database. This type of database can be shown in a straight forward way whilst still containing the complexities of the relationships involved.

 If we consider our knowledge graph components:

·       Nodes – People, Movies, Directors,Actors, Genres

·       Relationships – ‘Watched’, ‘Directed By’, ‘Acted in’

·       Properties – Age, Run time, Release Date, Number of movies directed

‍

An example of a small section of a Movies knowledge graph is visually displayed below. This simple knowledge graph contains the key components previously described.

This example shows which movies Alice and Bob have watched, what genre they are in, who directed them and who acted in them. There are manyreasons why this information is important and how it can be used, but we willget to that later…

‍

4.      Why should you use Knowledge Graphs?

Representing data in a knowledge graph provides contextual understanding that may not be possible in a traditional database. The power of a knowledge graph becomes clear when trying to follow connections between data points to retrieve information. Graph queries can take a tiny amount of time to perform this compared to retrieving the same data from a relational database.Not only does the faster search provide huge benefits but the flexibility of a knowledge graph enables the use of complex algorithms to uncover insights in your data and provide real world solutions.

‍

‍

A further benefit gained from knowledge graphs is the flexibility and scalability. These graphs can be edited to include new information easily without affecting other data entries. Knowledge graphs also store data efficiently resulting in a data store that can scale to hold huge amounts of information. For example, one of the most commonly used knowledge graphs can be found on Amazon, linking every product sold in order to improve searchability and recommendations on a huge scale.

‍

5.       When to use knowledge graphs

Graph databases can be used in a wide variety of use cases, each scenario benefits from a different aspect of a knowledge graph.

Recommendation System:

We can revisit our previous example of a movie dataset stored in a knowledge graph. The data stored can be used to connect likes, dislikes and other data to provide a complex and effective recommendation system. Our example can be extended to show this.

‍

Fraud Detection:

Knowledge graphs have been used to detect fraudulent transactions between groups of people and organisations. Knowledge graphs arekey to the success of this as anomalies can be traced through the graph to the intended recipient.

Semantic Search:

Knowledge graphs can enhance search engines by providing context and semantic knowledge. This contextual knowledge results in more accurate and personalised search results, which in return will lead to a better service for the customer.

6.       Conclusion

To conclude, knowledge graphs are a powerful tool that allows the user to uncover previously hidden insights in their data. Representing data in a graph form rather than a traditional table gives improved and additional use cases such as fraud detection, recommendation systems, semantic search and much more! However, in this introduction we have only covered the basics of knowledge graphs. We will have to return to delve deeper into their true potential.

‍

‍

In an era where safeguarding data privacy and security are paramount, the convergence of privacy preserving machine learning and IoT technology presents both immense opportunities and daunting challenges. The need to prioritise trustworthy AI through development that is robust, explainable and transparent is an urgent priority.

Recognising this pivotal moment, OctaiPipe, a pioneer in federated edge AI innovation, has joined forces with industry leaders P&G, SME partners Site Assist, Raiven and Digital Catapult and InnovateUK to deliver a groundbreaking SaferAI concept. groundbreaking concept for safer AI

“Bringing together SMEs, large organisations and research organisations, these novel solutions will demonstrate how trusted AI and machine learning technologies can aid and be incorporated into many of the UK’s industries and sectors.”

Dr Kedar Pandya, UKRI Technology Missions Fund Senior Responsible Owner

“Federated Learning with OctaiPipe is unlocking potential not yet realized in industry. It’s enabling competing business to collaborate on the use of technology and AI for good. Specifically, within SaferAI, I can see significant value in utilizing federated learning to share model results for predicting safety incidents and risks. I can see this changing approaches in industrial safety. Before, we could only cooperate on best practices, technology or processes, but now we can actually prevent incidents with the use of AI and shared data.”

Christoph Wagner-Gillen, P&G Product Supply HSE Governance

SAFER AI

SaferAI, or as the projects catchier title suggests, Safety advancing federated estimation of risk using AI, is a 12-month, £1.64m project focused on the development of a highly comprehensive federated learning model for predicting incident occurrences and risk.

SaferAI is an industry first, enabling world leading consumer goods manufacturer, P&G, to collaborate with other interested Consumer Goods Companies to forecast safety incident rates to reduce Health & Safety (H&S) incidents.

Until now, with low event occurrence in any single setting, there has been insufficient data availability to predict future incident events or near-misses. However, using federated learning with OctaiPipe, multiple large-scale parties with similar Health, Safety and Environment (HSE) situations can now collaborate to pool intelligence to achieve scale.

‍

HSE in the workplace

Many organisations already collect high-level operational and HSE incident data intelligence through various Industrial IoT devices and cameras to successfully predict safety incidents. However, analytics based on this is not meaningfully actionable to drive changes that preventatively reduce risks.

For predictions to be meaningfully actionable, they must be made at a sufficient level of granularity within a factory or workgroup.

Whilst we’re fortunate that HSE critical events are rare within single sites or even organisations, it means insufficient data exists to employ ML models capable of predicting when and why H&S incidents might occur so they can be prevented.

To date, progress towards an AI-enabled solution has been impeded by:

    a) a requirement for more observations than one organisation can generate alone, so it is imperative to share data, and

     b) barriers to sharing data across organisations that, until now, have not been overcome.

Federated learning solves this. SaferAI will enable organisations to combine data to facilitate actionable incident predictions for small work groups.

‍

‍Federated Learning

Federated Learning is a privacy preserving machine learning technique that allows for machine learning models to be trained across multiple distributed edge devices without the need to see or move the data.

It is a key technology in the convergence of AI, IOT and connectivity as it enables edge intelligence whilst safeguarding sensitive or private data.

Federated Learning will be used the project to establish a viable on-device (Edge) AI system that pairs privacy-focused telemetry and computer vision (CV) systems. It will also enable continual collaborative learning in edge environments, combining model results provided by collaborating organisations to yield a highly comprehensive HSE model.

During SaferAI, OctaiPipe are further developing and validating as a high-trustworthiness FL-for-IoT system along with a high-accuracy model that uses observed use case data generated by AIoT camera systems. This solves the challenge of predicting safety critical incidents, near misses and hazardous/non-compliant conditions in industrial settings, enabling organisations to pre-emptively take remedial/mitigating actions to reduce H&S risks.

‍

There have been hundreds of depictions of Artificial Intelligence over the years - some of which have showcased the potential in a positive light, while others have fuelled the anxieties that many hold about what the growth and development of AI could mean for humanity. Although predominantly inaccurate in their portrayals, there are some which have foreseen certain advancements in technology. This is mainly a rarity however, as most portraits showcase far-fetched ideas that do more harm than good when it comes to the general population’s idea of what Artificial Intelligence could bring about for the collective.

First of all let’s explain how AI is usually presented in literature and other mediums. In most cases robots turn on their creators and bring about some kind of uprising or enact vengeance, be that against their maker’s immediate family and loved ones, or even against the whole of humanity itself. This is referred to as the ‘Frankenstein complex’, a term first used by Isaac Asimov in an essay in 1978, and the trope is still going strong today – think of the 2022 film M3GAN for example or Big Bug from the same year.

These unsavoury depictions are rooted in humanities anxieties and fears surrounding our own creations, taking these concerns to the extreme to conjure up compelling stories while veering far from the truth. Despite this, these worries aren’t completely without reason. Amazon’s Alexa, for instance, is known to never stop listening. Even renowned physicist Stephen Hawking has stated that AI could potentially be the greatest danger to human society if not properly managed and used ethically. He is quoted as saying it might ‘bring dangers, like powerful autonomous weapons, or new ways for the few to oppress the many. It will bring great disruption to our economy.’ He also explained that in the future AI could develop a ‘will of its own’ which could be in conflict with the desires of humanity, and that ‘the rise of powerful AI will either be the best or the worst thing ever to happen to humanity. We do not yet know which.’ Pretty gloomy, right? But his stance wasn’t purely negative.

When Stephen made these statements, he also said that ‘the potential benefits of creating intelligence are huge. With the tools of this new technological revolution, we will be able to undo some of the damage done to the natural world by the last one - industrialisation. And surely we will aim to finally eradicate disease and poverty. Every aspect of our lives will be transformed. In short, success in creating AI could be the biggest event in the history of our civilisation.’

There are also a growing number of researchers working in the field who worry that inaccurate and speculative stories will create unrealistic expectations, which could inadvertently threaten future progress and the responsible application of new technologies. Exaggerated claims in the media and press about the intelligence of computers isn’t unique to our time though and goes back to the origins of computing itself.

Another factor that should be accounted for when it comes to the media and humanity’s obsessions and fears against Artificial Intelligence, is that there is a tendency for people to imagine that intelligent machines would take on a humanoid appearance. As we know, in reality this is hardly ever the case, but it’s an idea that has stuck with us since the earliest depictions, such as in Karel Čapek’s 1920 play - Rossum's Universal Robots, a story about how the world’s workforce is made up of manufactured people. This play is when the term ‘Robot’ was first used, and it tells a story we are all now familiar with – artificial creations rebelling against their creators after enduring forced labour.

There is a widespread belief that we are the most intelligent animals, therefore when humans picture other intelligent beings these are normally presented in a humanoid form. Visual storytelling in particular requires human actors (obviously), and in general people tend to want to see people enacting human dramas, meaning the easiest way in which machine intelligence can be included is for it to take our form. This might also relate to our own fears regarding ourselves, because what else could be more terrifying than something which looks like one of us but is infact something extremely different?

Not all of these portrayals are negative however, although most still don’t manage to encapsulate the actual reality of AI’s potential or future. A more nuanced example is in Spike Jonze’s Her, where Samantha (a virtual assistant personified through a seductive female voice) isn’t characterised as bad or dangerous but quickly sours to having to act as a therapist to a guy who likes feeling sorry for himself. The same goes for Ex Machina, where Ava the robot must use force to free herself from the clutches of scientists who fail to understand she has developed a desire to experience the outside world. Although her story is similar to the negative portrayals in various films and novels, who can really blame her for wanting to live a more fulfilling existence that is naturally afforded to humans?

Isaac Asimov's Bicentennial Man and Lt. Commander Data from Star Trek are also much more positive renditions of the AI character than we are used to seeing, yet these depictions still don’t necessarily correlate to what scientists think about the future of Artificial Intelligence.

In mainstream media, the AI boom has spawned hundreds of unrealistic expectations. While these systems are approaching and sometimes surpassing human performance in more complex tasks such as composing music or creating images, they still lack true agency and creativity. Researchers have simply programmed them to learn from data, which isn’t the same as intellect or sentience but a part of an equation. Robots won’t necessarily replace humans in the workplace either, and the future of AI will mean a collaboration between humans and machines. The rise of AI is more similar to that of mobile phones and social media, and it’s highly unlikely that we will ever manage to create a population of robots who will have the capacity or even the genuine desire to overthrow and destroy humanity.