NEC Quantum Computing Technology - ApplicationsNEC Fielding, Ltd.
By applying quantum computing to delivery plans, a two-hour task by experienced staff is shortened to a 12-minute automated operation.
- North America
Key point(s) of application
Issues and background
- Destinations and order of delivery: Delivery plan for service parts was overly complex and reliant on experienced staff
- Combinations of conditions are massive: Calculation would take too long to develop a delivery plan using AI or supercomputers alone
- Experienced staff is a valuable asset that can also be useful in operations other than delivery planning
Reduces delivery planning time
Quantum computing enabled the automatic generation of delivery plans, whose standard of accuracy reached the same level as that achieved by experienced staff, thereby attaining a significant streamlining of operations. Time spent on planning for delivery requests received by the day before has been shortened from 2 hours to 12 minutes.
Introduction effect can be expanded to a maximum of 30% increase in operational efficiency
Leveraging the mass, high-speed computational capability of quantum computing, up to 30% increase in business efficiency can be expected in the future by further repeating tuning and expanding the scope so that urgent orders can also be covered.
More active roles for the limited experienced staff
The automation of planning has enabled valuable experienced staff to play active roles in other operations. Further improvement in the value of operation and maintenance services can be expected.
Combinations of delivery routes that meet service parts demand with complex conditions in maintenance service amount to the 753rd power of 10 on a daily basis. Delivery requests received by the day before used to be planned by experienced staff based on their discretion, a process that would take two hours to do by hand. The annealing machine introduced as a solution shortened the time taken for delivery planning, of the same standard as that produced by experienced staff, to 12 minutes.
Pre-existing issues and background
Exploring ways that do not rely on the "specialist skills" of experienced staff
As a member of the NEC Group, NEC Fielding provides operation and maintenance services for servers, medical equipment, factory systems, and other machinery, covering a wide range of both ICT and non-ICT equipment. With round-the-clock services available 365 days a year nationwide, our network of about 360 bases supports the utilization of our customers' equipment and their business continuity.
Operation and maintenance services can be categorized into scheduled regular maintenance and troubleshooting for malfunctions of equipment and other problems. The latter is an emergency situation. We need to respond to these irregular occurrences flexibly and swiftly.
To build up our response capacity, we at NEC Fielding has taken on various challenges over the years. However, there were still issues we had yet to overcome. What stood in the way was the dependence on the "specialist skills" of experienced staff. Once we get a request for dealing with equipment failure, we arrange for the delivery of service parts and collection of failed parts while coordinating dispatch of engineers at the same time. The kind and location of malfunctions differ every day, so which service parts to load into the same vehicle, what route to take, and in which order to deliver the parts must be planned as such requests arise. Taking into account all aspects of delivery, including the required delivery time in synchronization with the arrival of engineers, hundreds of delivery locations, near-150,000 part types, and modes of transportation (light road vehicle, motorcycle, etc.), the possible delivery plan combinations can be as many as the 753rd power of 10.
Development of delivery plan takes too much time even when artificial intelligence (AI) or a supercomputer is used.
"Experienced staff has been attending to the massive combinations and making optimal delivery plans a day before the delivery based on experience and intuition such as 'this route takes less time even when the distance is longer' and 'this road is congested on these days of the week.' This reason made training successors for this operation extremely difficult. We were trying to find a way that does not rely on the specialist skills of experienced staff," says Mr. Masashi Yamazaki of NEC Fielding.
Considerations for selection
Seeing a way out with quantum computing
What NEC Fielding used as solution is quantum computing. Quantum computing has been talked about a lot in recent years. Its prominent feature is its outstanding computational capabilities. Processing that is said to take immense amount of time even with a supercomputer when simply working with computing power can be executed with hyper-speed in an unconventional way. The annealing machines, which we are already beginning to see in practical use, are great at deriving optimal combinations from among a multitude of possible combinations. It was indeed the perfect technology to solve the delivery planning issue that NEC Fielding was facing.
"When we were struggling to find an ideal way, we learned that annealing machines may be able to solve our problem. In order to solve the many issues the logistics industry is faced with, including driver shortages, surging costs, and carbon neutrality requirements, we wanted to check out the practicality of this technology. With that in mind, we decided to take on this challenge together with NEC," says Mr. Yamazaki.
Two-hour delivery planning shortened to 12 minutes
The utilization of the annealing machine enabled NEC Fielding to automatically generate delivery plans, which significantly reduced the time spent on this task. "We were able to reduce what previously took two hours to only 12 minutes. Experienced staff who used to make these delivery plans checks the generated plans and has confirmed that these are mostly of the same quality as the ones they would have created. Some generated delivery plans even came as a new discovery—a route suggestion that we wouldn't have noticed," says Mr. Yamazaki.
Further tuning and expansion of the scope to cover not only delivery planning a day before, but also planning for urgent delivery requests that arose on the day, are expected to further reduce the number of operating vehicles, driving distances, and costs, thereby achieving a maximum of 30% improvement in efficiency. Naturally, this will also contribute to reductions in CO2 emissions and environmental burden.
"We have repeatedly interviewed experienced staff, and the annealing machine incorporates their rich experience and intuitions in its computation conditions. For example, tacit knowledge dependent on individual skills such as road congestion trends and other such know-how was successfully transformed into codified knowledge. With this, we prepared and organized an environment to allow valuable experienced staff to perform what we actually wanted them to do aside from delivery planning. Taking advantage of the new system achieved with the annealing machine, we will work on further improving the value of our operation and maintenance services," says Mr. Yamazaki with enthusiasm.
Voice of NEC staff
Gaining a foothold in accelerating the practical application of quantum computing
Currently, NEC is actively engaged in efforts to develop and demonstrate practical use of quantum computers using quantum devices and quantum computing.
Besides the application of quantum computing in real quantum computers, quantum computing simulation software can be run on classical computers to enable these systems to emulate a quantum computer. The annealing machine used by NEC Fielding is an example.
Annealing machines can solve combinatorial optimization problems that were previously difficult to solve within reasonable time limits due to the vast number of combinations involved. Together with NEC Fielding, we were able to develop optimal delivery plans, optimize staffing according to the skill sets and job types of maintenance staff, efficiently load packages of different weights and sizes, and optimize delivery plans according to various conditions such as delivery time, work, and cost.
In working on this project, I also visited warehouses and observed the actual sites where the delivery plans were being implemented. What I sensed there was the wealth of experience, intuition, and above all, the grit of the seasoned employees. With the annealing machine, our goal at the beginning of the project was to incorporate as much of the rich experience of these employees as possible. NEC is very pleased to have achieved that goal and to see the potential for even greater efficiency gains. We will continue to expand our foothold in accelerating the practical application of quantum computing and using it to solve social issues.
NEC Fielding, Ltd.
|Head office address
|1-4-28 Mita, Minato-ku, Tokyo (Mita International Building)
|9,670.1 million JPY
|4,524 (as of end of March 2022, non-consolidated)
|Provides one-stop services that cover the entirety of the life cycle of an ICT system, from consulting to design, building, maintenance, and operation. Also provides installation, repair, and maintenance services for medical and nursing care equipment and other machines, as well as offers total support for customer systems.
(NOV 11, 2022)