1. Data silos: Data generated in the real world, is like the real world – it is imperfect. Eliminating silos and merging datasets together requires constant transformation, cleaning and structuring of that imperfect data. You can achieve this but with a manual approach would likely need hundreds of data analysts in your team. This is of course not feasible, so you learn to accept them. When you think of the multitude of data around security such as video cameras, access control, incidents, crime levels, threat intelligence asset registers, maintenance and guarding, how many different repositories does this data sit in? When you have silos you are making decisions about a 3D problem, not just in 2D, but from a limited slice of the 2D picture.
2. Data Analysis: Whatever data is analyzed is generally analyzed manually. Some things are very manual. A site audit literally requires a physical inspection to gather the requisite data. An auditor is of course limited by the time in the day, the travel to site, and so on. Then we have the semi-manual approaches of BI solutions. Even the most proficient BI teams and analysts teams cannot cope with the sheer magnitude of data around risk and supply chains.
3. Resilience: There is also the issue of resilience. Resilience requires foresight. We need to anticipate risks, strategically manage resources and budgets, and adapt and innovate to enhance capacity to withstand and recover from adverse events. But when you operate in the world of manual processes around your data, it is extremely challenging to have broad foresight. Siloed, manual processes produce a incomplete and static snapshot in time of a mere 2D sliver of our complex 3D environment which continues to dynamically evolve. Think about ensuring compliance with standards. The minute the auditor leaves, something might change on the site and the report is no longer valid. And what happens if a standard changes tomorrow, how long will it take to understand the impact? The insights you receive from your BI analysts is based on data that is already hours, days or weeks out of date. Data generated and analyzed by manual processes is always out of date, and this means decisions you are making about the future are not grounded in the world as it is at this specific moment. And to compound the problem, as humans we are not good at crunching millions of data points to understand trends and correlations in order to predict future trends and events.
When you deal with complexity at scale with manual processes around your data it is challenging to effectively deal with risk, protect your people, premises and assets, avoid erosion of your businesses profit margin, and create resilience.
The common definition of AI is the field of technologies that allows machines to execute tasks normally performed by humans. The reality is that AI can perform certain tasks that we would struggle to efficiently complete or cannot complete at all as humans.
If we return to the challenges and look at how AI can help:
1. First, in terms of silos, AI can automate the processes of mapping, transforming, cleaning and structuring data, making something that is otherwise time and resource prohibitive with manual process, suddenly achievable. Furthermore, unlike BI queries, there is more flexibility as the data does not always need to be consistent or perfectly structured –since deep learning models can understand data within context. Think of how many typos and errors there are when you ask ChatGPT a mangled question and it still understands what you meant. This flexibility suddenly allows data that was previously too dirty, to be suddenly usable and have value.
2. With manual data analysis AI excels in processing large volumes of data in real-time and performing the same tasks with consistency and without fatigue, unlike humans.
3. And with respect to prediction, deep learning models are adept at discovering hidden structures and patterns within millions of data points and uncovering complex, impactful and forward looking insights about your environment. AI systems also learn from data and improve over time without human intervention, adapting to new patterns and trends
AI offers you the ability to break down silos, leverage your imperfect data, process data in real-time, and start predicting future events. All at scale. All without more resources.
It is crucial to note that while AI excels in these areas, human intelligence remains unparalleled in many domains. Therefore, ignoring AI and its capabilities is a mistake, but only relying on AI is equally misguided. We need think of AI not as a replacement for us but as a complimentary technology that can step up our capabilities by empowering us to focus on more complex and creative endeavors.
