Infrastructure Monitoring for Predictive Maintenance – Updated 01/18/2019

We highlight the contributionsPonte Morandi, collapses are avoided with AI and IoT predictive maintenance of all infrastructures” (key4biz, October 3, 2018), “SHM systems for infrastructure maintenance and safety” (BitMAT ITIS Magazine, 22 October 2018) and “Crash-proof infrastructure, IoT and artificial intelligence “foresee” the risks“ (CorCom, 30 November 2018) And "Roads and civil engineering works, the keystone is predictive maintenance” (CorCom, 18 January 2019)..

They offer some preliminary considerations on a topic of the utmost topicality and relevance (in Italy there are tens or hundreds of thousands of suspended structures potentially at risk of collapse!), of enormous impact on daily life and the development of our Society, with recommendations to intervene extremely quickly on a national scale. Starting from the existing structures (“brownfield”), to then apply modern structural prevention design criteria (eg “digital twin”) and ICT monitoring to new buildings (“greenfield”).

In order to investigate the issue as a whole (number of structures, costs, times, ...), a Working group operational on the project "Infrastructure Monitoring for Predictive Maintenance" (IMPREM), trying to focus on some concrete "use cases" representative of real situations to develop a concrete propositional initiative at a political-institutional level to be extended with full knowledge of the facts to Government decision-makers. As States General of Innovation we want to place our specific skills and professionalism in support of the Institutions that deem to take charge of the above, as regards advisory and technical support – including project or sub-project management – to the implementation of IoT / SHM monitoring (internet of things / structural health monitoring) and related prediction paradigms / Artificial Intelligence.

At the moment we have some initial ideas – partly derived from the contributions mentioned above.

1. An initial (rather crude) analysis would identify the annual allocation necessary for the 'predictive monitoring' of all suspended infrastructures in a fraction of a point of GDP.
2. Various contributions to the magazine genius they dealt extensively and with many articles on the collapse of the Morandi bridge starting from 15 August last, hoping for increasingly "resilient" infrastructures.
3. In this framework, i modern SHM systems (structural heath monitoring), which allow you to continuously monitor the state of health of the structures and maintenance targeted, preventing situations of collapse from occurring.
4. SHM systems allow you to check the health of the structure, through non-destructive evaluations(NDE) of the static or dynamic type, the localization of problems / defects, producing the necessary alarms / alert and indications on the residual life in safety (safe lifetime).
5. SHM requires multidisciplinary collaboration between Civil Engineering and Electronics / ICT, accelerating the transition from culture of project to that of the maintenance and management. It is not (only) a matter of positioning sensors / actuators and collecting / processing data, but of using the guidelines for structural monitoring (see UNI/TR 11634:2016 standard) also during the construction phase, e.g. with simulation techniques building information modeling (BIM).
6. A SHM system can monitor one or more structures, using both satellite observations in low orbit (for static type monitoring) and aerial observations (helicopters or drones) and from ground devices - which allow both static and dynamic monitoring, using sensors and very low consumption technologies (eg Time) connected via wireless sensor network (WSN) and the like.
7. Thanks to the wide availability of data collected in this way, it would be relatively easy to create one maintenance predictive ('on condition'), using the continued progress in (big) data mining / machine learning, with massive use of Artificial Intelligence (AI) to derive estimates on possible behaviors and measures to be taken in good time.

In summary, for the safety of the road network in operation, it would first of all be necessary to promulgate - in an extremely short time – binding framework rules to equate infrastructural prevention with the safety obligations established by current legislation. At the same time, activate an infrastructure mapping (See example of SINFI), building a computerized register of civil works (eg. blockchain-based) which identifies its age, structural parameters, state of "health", history of maintenance interventions, ... capable of constituting acadastre' organized by classes of potential risks and relative priorities of the maintenance interventions to be implemented, with contextual installation of adequate SHM sensors.

With regard to the civil works to be carried out ("greenfield"), impose the application of the aforementioned framework standards of infrastructural prevention, and the related modern techniques of structural design to the geological, environmental, logistic, ... characteristics of the reference territory, with extensive use of ICT and "digital twin" methodologies / Enterprise 4.0 (industrial internet of things, IIoT), very low latency communications (5G), AI algorithms and models for predicting risk events, etc.

Last but not least, a precise and constant information to citizens on the potential risks in the transitional period of making the entire national road network safe, which - far from creating panic situations - puts the population in a position to be able to implement all the precautions and the most appropriate route choices for the respective needs of safe travel .

«It is necessary to know in order to be able to decide» (L. Einaudi)