Dispatching field workers is not rocket science, but it is Quantum Physics

We live in a world increasingly driven by the ability to collect data from almost everything from human beings to physical infrastructures. The volumes and complexity of such data require more and more computational power to process, organize, extract information and support decision making to solve health, economic, environmental or any other kind of societal problems.

In the energy industry, particularly in the field of electricity distribution, the opportunities brought by the development of Quantum computing are starting to unlock new ways to make the grids more efficient, resilient and sustainable.

In very simple terms, harnessing the power of Quantum computers that operate exponentially faster than conventional computers, due to operating in qubits, can help tackling increasingly more complex problems and optimizing large volumes of activities in our business. Here’s why it matters!

Tackling exponential complexity in the electricity distribution business

Enel is the world largest international distribution system operator with more than 74 million end-users in 8 countries, therefore managing an outrageous quantity of interactions with customers, field activities and grid events. Every year, Enel Global Infrastructure & Network automatically dispatches more than 32 million jobs to tens of thousands of fieldworkers and contractors in 13 distribution companies within the Group. Every day (including weekends and bank holidays, because electricity never sleeps) we need to assign jobs to the crews as well as efficiently design a daily route.

This is a context very similar to the well-known Travelling Salesman Problem (TSP ), one of the most studied problem in Mathematics and Computer Science. Its formulation is simple: “given a set of cities on a map, what is the shortest possible route that allows a salesman to visit each city exactly once and returns to the origin city?”. As Enel Global Infrastructure & Networks, we have more than 20 thousand fieldworker teams (our “salesmen”) performing more than 90 thousand jobs (our “cities”) daily.

That’s why Enel has developed since several years its own heuristics to solve the job assignment problem within its own Work Force Management (WFM) solution that has been developed across all countries of presence; such a solution is efficiently operating worldwide allowing to deliver a continuously improving quality of service to our customers. But as we project to grow our base of served customers from about 74 to 90 million in the next 10 years, while at the same time aiming at increasing the quality of service by more than 60%, we need to rethink the jobs assignment problem and many other operational processes with a new mindset of exponential complexity and by searching even better solutions.

Hence, we have to perform a quantum leap in the way we are thinking our operations and… we took the “quantum” challenge. We are going beyond traditional computer science, testing concrete applications of Quantum Computing.

Concrete quantum computing applications

Since the beginning of Computer Science, after the II World War, we have been used to deal with a technology doubling its computational power (again, an exponential growth!) every 18 months (namely, Moore’s Law). So if on a given day we had 1 GHz processors, after less than a couple of years we could have a 2 GHz processor. Similarly, whenever we had a 1 GB memory, after 3 years we would have a 4 GB of memory, and so on.

Nowadays, after more than 50 years something is slowing down: the size of the transistors and their interconnection are reaching the nanometer size that is becoming comparable with the atom size, resulting in the fading out of the traditional electronics principles. What this means is that we are reaching a traditional electronics performance plateau and the quantum mechanics principles are coming into play.

The world’s biggest research centers, universities, tech companies are jointly working to implement this quantum computers both on hardware and software side. On hardware side, there are different implementations of the single qubits leveraging superconductors, photons and even quantized mechanical vibration energy . On the software side new languages as well as problem formulations are being developed to efficiently exploit the computational power of quantum computers and speed-up the optimization algorithms or new languages are being developed to directly program quantum circuits and processors .

But as those breakthroughs are still under development, a concrete and viable opportunity to test this new approach is represented by the new Quantum Inspired problem formulations. These have been introduced to allow problems to be efficiently solved (e.g., Quadratic Unconstrained Binary Optimization, QUBO) by powerful enough quantum computers of the future, but also compatible with currently available clusters of Graphical Processing Units (GPUs). 

The Enel Grids Quantum Experience and future outlook

Let us return to the Enel fieldworkers jobs dispatching problem and find out how quantum computers and quantum inspired formulations come together

At Enel, we developed Q-BEAT, a Quantum Inspired Optimization formulation of the fieldworkers dispatching that has been both tested on quantum computers and GPUs cluster. The very first results are outstanding: taking as a baseline the current heuristic used by Enel, Q-BEAT is delivering a reduction of more than 20% of travelling time with respect to working time, while overall execution time of the algorithm is sped-up of more than 18 times on a standard cluster of GPUs.

What can we achieve with actual quantum computers in the future? Our testing shows that by employing the right quantum computing power, we can get comparable results in terms of quality while speeding-up the execution time by more than 500x.

That’s why after rolling out this new Quantum Inspired Optimization for WFM and jobs dispatching across the whole Enel Global Infrastructure & Networks soon, we will expand this experience applying quantum inspired optimization to other problems ranging from grid planning to investments allocation.

Some concrete examples of the potential applications we are exploring include Network Analysis for Global Actions Ranking and overall grid simulation; Planning, similar to WFM dispatching but with a longer view; meter reading routes to optimize the activity where remotely controlled meters are not yet available; or the placement of temporary Power generators during emergency conditions.

These are some examples of Enel Grids capability to promptly identify complex problems with practical application and address them through the implementation of the extraordinary potential of quantum computing, resulting in a competitive edge in this field.

But the most important consequence of this is that we can get our crews faster on-site to repair outages, reduce the environmental footprint of our grid operations, and increase the reliability of our infrastructure in light of more extreme operating climate conditions. In short, it will allow us to increase the quality and reliability of services for all our customers making sure people can select electricity as the easiest choice for a more sustainable future.