Operation
❯
Automate operation - Robotization
❯Device control using AI consisting of cloud computing and embedded system
Large Companies
NLP - Text summarization
For:
Media Intelligence Analysts
Scope:
Using Abstractive Text Summarization to reduce analysis costs for media monitoring.
Goal:
Improve Operation Efficiency
Robotic prehension of objects
For:
Customers, 3 rd parties, end users, community
Scope:
Outputting the end effector velocity and rotation vector in response to the view
from a red green blue depth (RGB-D) camera located on a robot's wrist.
Goal:
Improve Operation Efficiency
Adaptable factory
For:
Component suppliers (sensors, actuators), machine builders, system integrators,
plant operators (manufacturer)
Scope:
(Semi-)Automatic change of a production systems capacities and capabilities
from a behavioural and physical point of view.
Goal:
Improve Operation Efficiency
Empowering autonomous flow meter control - reducing time taken for proving of meters
For:
Process industries; humans
Scope:
Calibration of control devices
Goal:
Other
Device control using AI consisting of cloud computing and embedded system
For:
Equipment users, manufacturers, distributors
Scope:
Learn the user's preferred temperature in each situation for the control of home
appliances (air conditioning equipment).
Goal:
Other
Next Century Workforce: Partnering humans & robots to drive efficiency & growth
For:
Financial advisors
Bank employees
Goal:
Improved Employee Efficiency
Powering remote drilling command centre
For:
Oil and gas upstream sector; environment, humans
Scope:
Oil and gas upstream (Deployed in 150 oil rigs and 2,5 billion+ data points each)
Goal:
Other
Order-controlled production
For:
Customer, producing companies, broker
Scope:
Automatic distribution of production jobs across dynamic supplier networks
Goal:
Other
Robotic task automation: insertion
For:
Incorrect AI system use; new security threats
Scope:
Robotic assembly
Goal:
Other
Robotic vision scene awareness
For:
Customers, 3 rd parties, end users, community
Scope:
Determining the environment the robot is in and which actions are available to
it.
Goal:
Improve Operation Efficiency
Value-based service
For:
Customer (product user), platform provider, service provider, product provider
Scope:
Process and status data from production and product use sources are the raw
materials for future business models and services.
Goal:
Other
Device control using AI consisting of cloud computing and embedded system
For:
Equipment users, manufacturers, distributorsGoal:
OtherProblem addressed
Keep rooms comfortable by running home appliances (air conditioning
equipment) at the user's preferred temperature according to the situation.
Scope of use case
Learn the user's preferred temperature in each situation for the control of home
appliances (air conditioning equipment).
Description
Motivation:
The temperature at which the user feels comfortable varies
depending on the conditions outside the air conditioner,
such as the outside temperature, intensity of sunshine, time
of day, day of the week, etc. Always maintain a comfortable
environment by eliminating the need for this setting change.
Problem statement:
Though the temperature at which the user feels comfortable
depends on the situation, such as the time of day and the day
of the week, it is impossible to present these settings at the
time of product shipment. Even if the designer of the product
provides a method to enable the user to set such a setting,
the user himself/herself does not know he/she is expected
to set at what degree on what time. Long-term data cannot
be stored in the device, so the model is forced to learn in the
cloud. Training the model in the cloud takes longer to be able
to cope with the sudden variations in the operating pattern
of the user.
Current situation: The temperature is set using the controller
every time the user feels uncomfortable.
Solution approach and solution steps:
In addition to training the model using long-term historical
data in the cloud, the model is also adjusted by frequent
training in embedded devices. When the user changes the
temperature setting using the controller, not only the setting
but also accompanying data, such as the setting time, are
stored in the air conditioner. Data about the operating status,
such as temperature sensor values installed for the control
of the air conditioner, are stored in the air conditioner. Data
stored in the air conditioner is periodically uploaded to the
cloud instance held by the manufacturer. The latest weather
forecast information, etc. is kept on the cloud service at all
times. A model is created to represent what the set
temperature is expected to be depending on the external
conditions around the air conditioner (including the
forecast) by periodical learning for each air conditioner on
the cloud. The model is delivered to the corresponding air
conditioner. Online machine learning is performed based on
the data stored inside the air conditioner, and the internal
parameters of the model are adjusted. This embedded
learning is performed frequently, e.g., once an hour, and it is
possible to reflect sudden changes in the user's usage pattern
to the model. The online machine learning algorithm inside
the air conditioner and the batch machine learning algorithm
in the cloud are tuned as close as possible to prohibit radical
change of the model from the model adjusted by online
machine learning when the model is delivered via cloud
computing and the adjusted model is overwritten. The air
conditioner predicts the preferred temperature by using the
model, and the result is used as the set temperature of the air
conditioner. As in normal operation, the air conditioner
performs control so that the temperature of the room
remains at the set temperature.
Results and effects:
Since the prediction is done by the air conditioner
(embedded), it keeps working even if there is a network
failure or a cloud failure. The only impact of a failure is the
inability to upload data and the inability to update the model
as it learns via the cloud. Operation training with a long-term
cycle, such as a fixed operation for each day of the week, is
effective if the model is trained from the accumulated
operation history. A model with this effect is created mainly
by learning on the cloud. If there are sudden operation
pattern changes, e.g., when the outside temperature rises
suddenly and the user reacts to it, high frequency online
machine learning inside the air conditioner can adjust the
model immediately.