TeamBrain free for 15 days
The subject of the smart city (‘Smartcity’) is often approached from the angle of connected objects (‘IOT’) (e.g. streetlight, VMC, touch road, on-board sensors …). In this context, the citizen becomes a source of information through his behaviors and the data he generates. This is a precursor to tremendous progress: public and private actors can analyse, or even predict, the needs of citizens and thus adapt their policies and services. This Smartcity could be described as ‘top-down’.
However, despite the fact that the citizen is the subject of all the statistical attention, he is paradoxically badly, if not not ‘listened to’. Wouldn’t the challenge facing local elected officials be to be able to provide immediate, personalized and ‘certified’ answers by showing active listening, while relieving the (human) referents of its Administration?
Smartcity fairs, in which we participate, are often oriented around connected objects, energy, the exploitation and optimization of resources, natural resources, improvement of the environment. This development is structural and commendable.
Similarly, many application maps published by Smartcity specialists exist. They list the components and applications in the smart city ecosystem, without however addressing the subject of active citizen listening.
Indeed, it is surprising that the Smartcity is rarely considered from the angle of optimizing relationships and information shared with the citizen. However, it seems to us that the notions ‘citizen 3.0’, ‘administration 3.0’ should take a greater place.
Artificial intelligence cannot, at present, completely reliably answer citizens’ questions. Unlike image recognition, so-called ‘artificial intelligence’ software still misunderstands and struggles to remove ambiguities in language. Artificial intelligence, as powerful as it is, is not able to understand and interpret the provisions of the Civil Code or the Urban Planning Code for example (…).
At most, it would be able to orient towards a text, such as an open-text search engine… but it does not fulfill the purpose of listening, understanding and answering specific questions.
However, citizens should be able to formulate their questions and receive precise and ‘certified’ answers.
The main tools on the market are those of the ticketing type (1 question = 1 ticket), FAQ (frequently asked questions) more or less elaborate, open-text search engines (see Google), or chatbots. However, these tools do not meet the objective or involve heavy investments (30,000 to +150,000 € per process), to ultimately obtain a ratio of 50% satisfaction (60-70% in some specific cases).
In our opinion, the most efficient and economical solution is the use of a dynamic FAQ associated with administrative teams: using hybrid solutions that combine human and machine learning.
Machine learning, combined with human referents, makes it possible to take advantage of the best of both worlds: the machine is perfect for automatically processing redundancies and applying workflows; the human thus ‘augmented’ finds all its value in knowing how to interpret complex rules, make links with related subjects, all while appealing to a critical mind and a judgment imbued with hindsight and sensitivity. The human singularity would become an asset in a world invaded by robots.
Public and private actors could listen to citizens, respond effectively and continuously improve thanks to their ‘augmented agents’. In perpetual search for economy and efficiency, they could optimize their services quantitatively (fewer agents to deal with redundant questions or not) and qualitatively (better listening and alignment of answers), for the happiness of users, customers and administrators.