Artificial Intelligence for scheduling field operations

How can Artificial Intelligence transform Field Service Management?

Artificial intelligence is a discipline that studies the theoretical foundations and methodologies that enable the design of computer systems to behave with human-like intelligence.

Artificial intelligence techniques include:

- Machine learning - helping machines to learn from experience through the application of statistical methods to data;
- Deep learning - uses neural networks to learn from data.

For some years now, artificial intelligence has been playing a leading role in the market, mainly due to the following four conditions:

- Ability to tap into very efficient algorithms;
- Availability of powerful computing within company data centers or the cloud;
- Willingness of companies and Governments to invest in these technologies;
- Spread of applications capable of generating the data needed for artificial intelligence.

Terranova’s goal has always been to provide customers with tomorrow’s technology, today.
This resulted in the strategic acquisition of HPA, a spin-off from the University of Verona that develops customised solutions for predictive analysis in the utility sector, using machine learning and deep learning techniques.

Of the many projects now underway, the application of artificial intelligence can be analysed in the context of optimising the scheduling of field operations.

The use of artificial intelligence within Terranova’s applications has resulted in:

- dramatically improved execution times;
- optimal distribution of activities and better organisation of operators' work;
- comparison of different management situations to choose the best solution;
- reduction in kilometres covered by the vehicle and related decrease in fuel consumption and CO2 emissions;
- proactive reorganisation following unforeseen events using automatic planning that is communicated to all operators involved;
-ability to manage seasonality and hence greater awareness of the behaviour of operators and the environment in general at all times of the year.

To optimise field operations, however, it is not enough to adopt an algorithm that just predicts the average duration of operations. It is also important to include location, season or operator in order to avoid mistakes.

Instead, it is essential to have information about the activities to be performed, the skills of the field service technician, the equipment at their disposal and the territory of operation. This subsequently allows the best combination of these elements in order to achieve better results at lower cost.

Field-based operational excellence essentially depends on a combinations of optimisation algorithms and robust datasets. This allows the identification of the best and most suitable field service technician for each task type as well as an indication of how long they will take to complete the activity.

The Terranova Work Force Management solution (TWFM) allows the adoption of the following maintenance strategies:

- proactive maintenance: periodic interventions designed to anticipate and eliminate asset failures;
- reactive maintenance: interventions are carried out in direct response to identification of asset fault/failure;
- predictive maintenance: interventions based upon informed predictions which accurately identify future asset failure/fault according to predefined thresholds being exceeded.

According to a study by the Deloitte's Analytics Institute[1], predictive maintenance can increase company productivity by 25%, reduce breakdowns by 70%, and cut maintenance costs by 25%. Terranova can help improve on these metrics through the use of an unique algorithm that can be applied to both predictive and reactive maintenance regimes.

This is made possible by the continuous availability of real -time data to enable more informed and focused interventions by field-based teams well before faults occurr and negatively impact company operations.

Therefore, this collaboration between artificial intelligence and field-based technicians presents a compelling leap forward in field service management.

Terranova pursuit of using artificial intelligence within its software to improve field activity efficiencies went even further with modifications in areas of image recognition and mobile vocal assistance.

IMAGE RECOGNITION

Field-based operatives take photographs as part of routine meter reading activities and refer back to these images on the rare occasion that meter readings are contested. However, these photographs are also a massively under utilised and valuable data source that AI can rapidly catalogue and extract meter readings, meter models, and more.

The goals of this project are:

- Scalability - for ultra quick background process that enables swift and automatic analysis of new and existing images;
- Data quality - to exploit under utilised photographs taken as part of routine meter reading activities. These untapped images are a valuable data source;
- Expandability - this project can easily be extended to cover existing and future meter models;
- Continuous improvement - the engine can achieve near 100% accuracy as it rapidly learns from operator lead corrections that lower error margins.