Automating factory inspections with temi: FieldLog use case study
This document explains the specific setup method and operational flow for automated patrols using temi and FieldLog in conjunction.


We can't allocate personnel for nighttime and holiday inspections of the factory. However, we can't overlook equipment malfunctions—the solution to this dilemma is...Automated patrol by autonomous robots.
This article provides a detailed explanation of the mechanisms behind robotic technology that enables reduced manpower for nighttime patrols, and the specific effects of its implementation.
Nighttime and weekend inspections of manufacturing sites present a common challenge for many factories.

Securing staff for night shifts and weekend work is becoming increasingly difficult each year. This is especially true for factories in rural areas, where recruiting young people is challenging, and the existing staff is aging."Even with increased night shift allowances, we're not getting any applicants." "We're overly reliant on veteran employees, leading to a highly dependent workforce."Such comments are not uncommon.
Amidst a worsening labor shortage, reducing the number of patrol personnel from two to one at night has resulted in a decline in the quality of inspections in some cases. A one-person system reduces the ability to respond to emergencies and increases the risk of accidents.
The quality of manual inspections varies greatly depending on the skills and experience of the person in charge.Experienced employees can detect subtle abnormal noises or changes in vibration in the equipment, but less experienced staff tend to overlook them.
Inspection routes can vary slightly depending on the person in charge, and there have been cases where abnormalities were not detected because "they weren't looking in places different from usual." Furthermore, maintaining concentration is difficult during late-night hours, and the risk of overlooking things due to fatigue cannot be ignored.
With handwritten inspection sheets, the record of "who inspected what and where" tends to be subject to individual differences. When the person in charge changes, it becomes difficult to read the records, it takes time to find past records, and managers cannot grasp the situation in real time unless they are on-site—this "invisibility of records" leads to inefficiencies on the ground.
Furthermore, from the perspective of ISO certification and quality management, there is a growing demand for the digitalization of inspection records."To make visible who inspected what."Starting from there will be the first step towards creating a future record management system.

Reducing manpower in nighttime patrols means,Replacing nighttime and holiday equipment inspection and patrol tasks, which were previously performed by humans, with autonomous robots.The robot automatically patrols a set route and photographs and records equipment and instruments at scheduled times.
Temi, an AI-powered autonomous robot, uses LiDAR sensors and cameras to create a map of the factory using SLAM technology, and automatically navigates to its destination while avoiding obstacles. It can travel at a maximum speed of 1 m/s and adjust its speed according to the site conditions. It can also pass through narrow passages (minimum width of 60 cm).

It operates continuously for approximately 8 hours on battery power and automatically returns to the charging station when charging is needed. For factories requiring 24-hour operation, a system of multiple units taking turns charging can be established.
[Regarding the lighting environment] For image capture and instrument checks during automated nighttime patrols, a certain level of illumination is required within the facility. While many manufacturing facilities have safety lights and night lights that ensure a minimum level of illumination at night, please consult us in advance if your lighting environment is limited, such as in areas where lights are automatically switched off by motion sensors.
While the robot alone can only "move," by linking it with a dedicated patrol inspection app, the following functions become possible.
What is FieldLog? This is an inspection record management app for manufacturing sites developed by iPresence Co., Ltd. It consists of two versions: a browser version and a version equipped with temi. Through temi's autonomous patrol and built-in camera photography, patrols, photography, and recording are digitized in one go.

The patrol will automatically start at a set schedule, such as 22 PM, 2 AM, and 6 AM every day.
The system stops in front of each piece of equipment and automatically takes pictures of the instruments and equipment status (recording the date, time, and location of the photo). OCR reading is also possible to acquire meter readings.
The captured data is uploaded to the cloud. The system allows you to check who inspected what and where.
If the set threshold is exceeded, an email notification is sent to the administrator. Images on the cloud can be viewed remotely, and temi can be remotely controlled as needed.
You can check the current location, view real-time video, and remotely instruct on-site personnel to perform additional inspections from your smartphone or PC.

Fixed cameras can only monitor the area where they are installed, making them prone to blind spots, and installation work is required. Similarly, IoT sensors are fixed in place, and the cost increases with the number of sensors.
Mobile robots can cover a wide area with just one unit, and their routes can be flexibly changed.Even if new equipment is added, it can be handled simply by adding checkpoints on the app. In addition to video monitoring, remote communication with on-site staff is also possible.

This guide explains the specific setup methods and operational flow for temi×FieldLog, from route configuration to automated patrols and cloud recording.
Let's consider a factory that had two people assigned to nighttime patrols, and estimate the actual implementation costs.
[Notes regarding pricing] The current Field Log license is ¥30,000 per month (excluding tax). We plan to revise the price to ¥50,000 per month when we expand the functionality (timing TBD). The price at the time of contract will be applied for 3 years. From the 4th year onward, the Field Log license fee will be the price applicable at the time of renewal.

* Estimated effectiveness may vary depending on the facility's lighting environment and the design of the patrol route.
The variability in inspections performed by humans will be eliminated.Inspections are always conducted along the same route, at the same checkpoints, and at the same time.Will be
Fixed-point photography allows for a time-series comparison of equipment conditions, which can be used for predictive maintenance. Changes such as "the area of rust has expanded compared to last month" or "instrument readings are gradually increasing" can be identified from the accumulated image data.
All inspection records are stored in the system, allowing for centralized digital management of "who inspected what and where":

[Digitizing records will change the workplace] By implementing FieldLog, inspection records that were previously buried on paper can now be instantly viewed on the system, showing "who inspected what and where." This eliminates the need to search through past records, and managers can grasp the situation on-site in real time from anywhere. This is a practical first step in steadily advancing the digitalization of the quality control system.
This eliminates the need to print, distribute, collect, and store paper checklists, contributing to a paperless office environment.
Currently, it is being implemented in multiple companies, primarily in the manufacturing sector.
The starting point was the need to remotely manage inspection work within the factory and to handle it at night. Existing water temperature gauges sound an alarm when the temperature exceeds a set value, but the challenge was to create a system that would not miss the "predictive zone" before that point. We began to consider whether the water temperature check, which was previously done visually by people, could be replaced by email notifications linked to a robot.
In this way, rather than replacing the existing alarm system,Add a mechanism to "notice before the alarm sounds."Complementary implementations are also possible. For details on implementation examples, please contact us and our representative will provide further information.
This involves replacing nighttime equipment inspection and patrol tasks, traditionally performed by humans, with autonomous robots. These robots automatically patrol pre-set routes, photographing and recording equipment and instruments at scheduled times. This not only reduces labor costs but also enables standardization of inspections and digitalization of records. Even in cases where complete automation is difficult, it's possible to reduce the number of personnel from two to one.
There are three main advantages.
① Staffing levels during nights and holidays can be reduced:If two people are assigned to the night shift, introducing robots can significantly reduce labor costs.
② Because inspection routes and times are fixed, quality is stable:This system eliminates human error and oversights, ensuring consistent quality inspections every time.
③ Centralized digital management of inspection records:The system allows you to check who inspected what and where, streamlining internal record-keeping and reporting processes.
For a 3-year lump-sum contract, the total cost for temi V3 SuccessPlan (3-year support) + Field Log license (3-year usage rights) is approximately 3.6 million yen (including tax). Compared to the personnel costs of two night patrol personnel (approximately 800,000 yen per month), this estimates that the investment can be recovered in approximately 4.5 months. From the 4th year onward, continued use is possible for approximately 900,000 yen per year (including tax, at current prices). Please note that while the price of the Field Log license is scheduled to be revised when features are expanded, the price at the time of contract will apply for the first three years.
Yes, we can accommodate that. The robot automatically creates a map of the factory upon installation and sets a patrol route and checkpoints on that map. It can travel on any path with a minimum width of 60cm and an incline of 5 degrees or less. The travel speed can be adjusted to suit the site, with a maximum speed of 1m/s, and the route can be flexibly edited via the app to accommodate layout changes.
While temi's autonomous driving itself can operate in dark places, a certain level of illumination is required for instrument capture and image recording using its camera. In most factories, sufficient illumination is provided by safety lights and night lights. If you are concerned about the lighting environment in each area, we recommend checking the site in advance.
FieldLog functions as a record-keeping platform that allows you to verify "who inspected what and where" within the system. By digitizing and centrally managing inspection records, it contributes to the efficiency of internal reporting and record organization. On the other hand, we are currently working on improving its functionality to meet the requirements for legal evidence (prevention of time tampering, long-term storage, etc.). Please feel free to contact us for specific ways in which it can be used in audits.
This article explains the FieldLog×temi implementation process, initial setup, and detailed costs.
You can also request a demo or consult about a Proof of Concept (PoC) from this page.
Replacing nighttime factory patrols with automated robotic patrols will yield the following benefits:
The initial setup cost was approximately 3.6 million yen (including tax) over three years, with the investment recovered in approximately 4.5 months. The cumulative difference over three years was approximately 25.2 million yen.
With fixed inspection routes and times, there are no oversights. Fixed-point photography also allows us to understand the changes in equipment over time.
Reduced risks associated with nighttime staffing; immediate notification and remote verification via alert in case of anomalies.
Reducing the burden of night shifts, alleviating staff shortages, and allowing staff to focus on value-added tasks.
Instead of "reducing inspections due to labor shortages," the approach should be "improving inspection quality by entrusting them to robots."This will become the standard for manufacturing sites going forward.
We recommend starting with a 1-2 month Proof of Concept (PoC) to verify whether it's suitable for your own workplace. It's being increasingly adopted by manufacturing companies, including metal surface treatment manufacturers.
* Estimated effectiveness may vary depending on the facility's lighting environment, patrol route design, and labor cost levels. Please feel free to contact us for details.

This document explains the specific setup method and operational flow for automated patrols using temi and FieldLog in conjunction.

This guide covers the entire implementation process, from initial setup to full-scale operation and how to apply for a PoC (Proof of Concept) rental.