Artificial intelligence at the service of “Facility Management” professions
Introduction
Artificial Intelligence regularly makes the headlines in the media. It is being implemented in a growing number of fields; however, many people are unaware of what AI really is and how it works. In this article, we recall the concepts of AI and its applications in the field of Facility Management (FM) which, faced with the health crisis, will undergo many changes.
What is Artificial Intelligence?
Artificial Intelligence (AI) is a concept which, due to its complexity, is difficult to define in a totally accurate way. According to Russell and Norvig, authors of the famous book “Artificial Intelligence: A Modern Approach”, AI is the study and design of intelligent agents; i.e. systems that can analyse their environments to make choices or take actions to maximise their performance for a given purpose. In other words, artificial intelligence is the ability of machines to “think” for themselves; thus, AI is demonstrated when a task, previously performed by a human and considered to require the ability to learn, reason and solve problems, can now be performed by a machine. Autonomous vehicles are an excellent example of this. There are generally two categories of AI: low AI and high AI.
Low AI is intended to solve a problem or make a decision within a well-defined framework. It encompasses almost all current applications of AI in fields as varied as natural language processing, recommendation systems, computer vision (digital recognition, face detection and recognition, etc.), medical imaging (detection of breast cancer by analysis of mammographic images, etc.), robotics, fraud detection, etc.
The concept of strong AI refers to an intelligence that could replace human intelligence in all its complexity. It corresponds to a machine capable not only of acting intelligently, but also of assimilating abstract concepts and having a real consciousness and feelings close to those experienced by human beings. Artificial intelligence researchers call the transition from weak to strong AI “singularity”. Singularity is considered to be the hypothetical upper turning point in technological evolution where artificial intelligence will supplant human beings. It has inspired science fiction films such as “Terminator”, “Matrix”, “Transcendence” etc.
AI works by generalisation on the principle that “everything that will happen has either already happened or is close to having happened”. It is based on machine learning, in particular deep learning and reinforcement learning, and uses big data to train models to reveal hidden paterns. The field of AI encompasses several other disciplines such as computer science, mathematics and statistics, natural language processing, psychology and neuroscience.
Artificial Intelligence and Facility Management
Many computer technologies have been adopted in recent years in the field of
Facility management (FM). These include CMMS management tools, intelligent sensors, cleaning robots and other technological applications. However, many often repetitive and time-consuming tasks that are still performed manually could be automated using AI tools. Machine learning algorithms coupled with natural language processing techniques can be used to build tools for automatic classification of service requests; while implementations of operational search and optimisation algorithms can help automate task planning, optimise resources or manage SLAs. In this context, benefiting from tools using Artificial Intelligence technologies can help address these issues more effectively.
Artificial Intelligence dedicated to the optimisation of Facility Management activities
There are various tools for FM actors but still too few of them embody artificial intelligence. The advantage of AI solutions is that they allow the creation of different modules dedicated to solving a given type of problem. Predictive analysis, intelligent planning, scheduling and optimisation or automatic natural language processing are all modules that would increase the competitiveness of FM players.
Predictive analysis modules, for example, use the most advanced Machine Learning algorithms to analyse and enrich data, identify and train models that best describe the available data and finally make simulations or predictions to enable decision-makers to make the best possible decisions.
The scheduling and constrained scheduling modules combine Operational Research and optimisation techniques with a rule-based system to address task allocation and scheduling problems in order to make the best use of resources while taking into account both task and resource constraints.
Finally, with an NLP module that essentially addresses the issues of sentiment analysis and document classification using deep learning algorithms, the ability to decipher an emergency situation in order to prioritise it while reorganising its planning in an optimal way becomes automatic and saves considerable time.
Of course, artificial intelligence is such a vast field that it would be presumptuous to present it in a few lines. This is a quick overview of concepts to illustrate how certain FM activities can be automated by taking advantage of AI, to the benefit of employees and the organisation.
By Adaté Tossa, R&D Engineer and Data scientist at DATAPOLE and Frédéric Gagnaire, Chief Executive Officer at DATAPOLE.