Controller and method using machine learning to optimize operations of a processing chain of a food factory
Computing device and method using machine learning to optimize operations of a processing chain of a food factory. The computing device collects data representative of characteristics of a product processed by the processing chain. At least some of the collected data are received from one or more sensor monitoring operations of the processing chain. The computing device determines at least one product characteristic value based on the collected data. The computing device executes the machine learning inference engine, which uses a predictive model for inferring command(s) for controlling processing appliance(s) of the processing chain based on inputs. The inputs comprise the at least one product characteristic value. The computing device transmits the command(s) to the processing appliance(s) of the processing chain. Examples of product characteristic values comprise: a product temperature, a product humidity level, a product geometric characteristic, a product weight, and a product defect measurement.
1. A computing device comprising:
at least one communication interface;
memory for storing a predictive model; and
a processing unit comprising one or more processor configured to:
collect data representative of characteristics of a product processed by a processing chain, at least some of the collected data being received via the at least one communication interface from one or more sensor monitoring operations of the processing chain;
determine at least one product characteristic value based on the collected data;
collect additional data, the additional data comprising at least one current quality metric, each current quality metric being defined for one of the processing chain or a processing appliance of the processing chain;
execute a machine learning inference engine, the machine learning inference engine using the predictive model for inferring one or more output based on inputs, the inputs comprising the at least one product characteristic value and the at least one current quality metric, the one or more output comprising one or more command for controlling at least one processing appliance of the processing chain; and
transmit via the at least one communication interface the one or more command to the at least one processing appliance of the processing chain.
2. The computing device of claim 1 , wherein the machine learning inference engine is a neural network inference engine implementing a neural network using the predictive model for inferring the one or more output based on the inputs, the predictive model comprising weights of the neural network.
3. The computing device of claim 2 , wherein the neural network comprises an input layer, followed by fully connected hidden layers, followed by an output layer; the input layer comprising at least one neuron receiving the at least one product characteristic value; the output layer comprising one or more neuron outputting the one or more command; the weights of the neural network being applied to the fully connected hidden layers.
4. The computing device of claim 1 , wherein the processing chain is located in a food factory and the product is a food product.
5. The computing device of claim 1 , wherein the one or more command controls a functionality implemented by the at least one processing appliance, the functionality comprising inspecting, sorting, cleaning, cutting, peeling, slicing, blending, mixing, blanching, cooking, baking, frying, heating, cooling, freezing, humidifying or packaging.
6. The computing device of claim 1 , wherein at least some of the collected data representative of characteristics of the product processed by the processing chain are received via the at least one communication interface from one or more information server.
7. The computing device of claim 1 , wherein the at least one product characteristic value comprises at least one of the following: a temperature of the product, a humidity level of the product, a geometric characteristic of the product, a weight of the product, a tensile strength of the product, an internal pressure of the product, a stock keeping unit (SKU) of the product, and a defect measurement for the product.
8. The computing device of claim 1 , wherein the additional data further comprise at least one environmental characteristic value; the at least one environmental characteristic value comprising at least one of a temperature of an area where the processing chain is located, a humidity level of the area where the processing chain is located and a lighting level of the area where the processing chain is located; the inputs of the machine learning inference engine further comprising the at least one environmental characteristic value.
9. The computing device of claim 1 , wherein the additional data further comprise at least one current operating parameter of at least one processing appliance of the processing chain; the inputs of the machine learning inference engine further comprising the at least one current operating parameter.
10. The computing device of claim 1 , wherein the additional data further comprise at least one target yield; each target yield being defined for one of the processing chain or a processing appliance of the processing chain; the inputs of the machine learning inference engine further comprising the at least one target yield.
11. The computing device of claim 1 , wherein the additional data further comprise at least one current yield; each current yield being defined for one of the processing chain or a processing appliance of the processing chain; the inputs of the machine learning inference engine further comprising the at least one current yield.
12. The computing device of claim 1 , wherein the additional data further comprise at least one target quality metric; each target quality metric being defined for one of the processing chain or a processing appliance of the processing chain; the inputs of the machine learning inference engine further comprising the at least one target quality metric.
13. The computing device of claim 1 , wherein the additional data further comprise at least one target carbon dioxide (CO 2 ) footprint; each target CO 2 footprint being defined for one of the processing chain or a processing appliance of the processing chain; the inputs of the machine learning inference engine further comprising the at least one target CO 2 footprint.
14. The computing device of claim 1 , wherein the additional data further comprise at least one current CO 2 footprint; each current CO 2 footprint being defined for one of the processing chain or a processing appliance of the processing chain; the inputs of the machine learning inference engine further comprising the at least one current CO 2 footprint.
15. The computing device of claim 1 , wherein the determination of one or more among the at least one product characteristic value based on the collected data uses another machine learning inference engine.
16. A method using machine learning to optimize operations of a processing chain, the method comprising:
storing a predictive model in a memory of a computing device;
collecting by a processing unit of the computing device data representative of characteristics of a product processed by the processing chain, at least some of the collected data being received via at least one communication interface of the computing device from one or more sensor monitoring operations of the processing chain;
determining by the processing unit of the computing device at least one product characteristic value based on the collected data;
collecting by the processing unit of the computing device additional data, the additional data comprising at least one current quality metric, each current quality metric being defined for one of the processing chain or a processing appliance of the processing chain;
executing by the processing unit of the computing device a machine learning inference engine, the machine learning inference engine using the predictive model for inferring one or more output based on inputs, the inputs comprising the at least one product characteristic value and the at least one current quality metric, the one or more output comprising one or more command for controlling at least one processing appliance of the processing chain; and
transmitting by the processing unit of the computing device via the at least one communication interface of the computing device the one or more command to the at least one processing appliance of the processing chain.
17. The method of claim 16 , wherein the machine learning inference engine is a neural network inference engine implementing a neural network using the predictive model for inferring the one or more output based on the inputs, the predictive model comprising weights of the neural network.
18. The method of claim 16 , wherein the processing chain is located in a food factory and the product is a food product.
19. The method of claim 16 , wherein the at least one product characteristic value comprises at least one of the following: a temperature of the product, a humidity level of the product, a geometric characteristic of the product, a weight of the product, a tensile strength of the product, an internal pressure of the product, a stock keeping unit (SKU) of the product, and a defect measurement for the product.
20. The method of claim 16 , wherein the determination of one or more among the at least one product characteristic value based on the collected data uses another machine learning inference engine.
21. A non-transitory computer-readable medium comprising instructions executable by a processing unit of a computing device, the execution of the instructions by the processing unit providing for using machine learning to optimize operations of a processing chain by:
storing a predictive model in a memory of the computing device;
collecting data representative of characteristics of a product processed by the processing chain, at least some of the collected data being received via at least one communication interface of the computing device from one or more sensor monitoring operations of the processing chain;
determining at least one product characteristic value based on the collected data;
collecting additional data, the additional data comprising at least one current quality metric, each current quality metric being defined for one of the processing chain or a processing appliance of the processing chain;
executing a machine learning inference engine, the machine learning inference engine using the predictive model for inferring one or more output based on inputs, the inputs comprising the at least one product characteristic value and the at least one current quality metric, the one or more output comprising one or more command for controlling at least one processing appliance of the processing chain; and
transmitting via the at least one communication interface of the computing device the one or more command to the at least one processing appliance of the processing chain.