Web services platform with cloud-based feedback control

1 . A web platform comprising:

one or more processors; and

one or more storage media storing instructions thereon that, when executed by the one or more processors, cause the one or more processors to:

receive a timeseries of data values from a piece of equipment of an environment located remote from the web platform;

lookup, using a timeseries identifier of the timeseries of data values, a control workflow from a database storing a plurality of control workflows for different pieces of equipment, the control workflow uses the timeseries of data values as an input and defines one or more processing operations to be applied to the timeseries of data values responsive to a detection that the control workflow executes for the piece of equipment;

execute the control workflow based on the timeseries of data values to generate a control signal; and

provide the control signal to the piece of equipment of the environment, the piece of equipment operating based on the control signal.

2 . The web platform of claim 1 , wherein the instructions cause the one or more processors to:

store the database storing the plurality of control workflows comprising the control workflow on the one or more storage media, the control workflow of the plurality of control workflows to execute for the piece of equipment and a second control workflow of the plurality of control workflows to execute for a second piece of equipment; and

select, from the plurality of control workflows, the control workflow that uses the timeseries of data values as the input.

3 . The web platform of claim 1 , wherein the timeseries of data values include a plurality of feedback values measured by a sensor of the piece of equipment of the environment.

4 . The web platform of claim 1 , wherein the instructions cause the one or more processors to:

identify one or more other timeseries required as inputs to the control workflow, wherein the one or more other timeseries comprise a setpoint timeseries comprising a plurality of setpoint samples, the plurality of setpoint samples defining setpoints corresponding to samples of the timeseries of data values; and

generate an enriched control workflow comprising the control workflow, the timeseries of data values, and the one or more other timeseries.

5 . The web platform of claim 1 , wherein the database is a directed acyclic graph (DAG) database storing a plurality of DAGs, each of the plurality of DAGs defining a particular control workflow;

wherein the instructions cause the one or more processors to:

determine whether any of the plurality of DAGs stored in the DAG database are configured to use the timeseries of data values as a DAG input.

6 . The web platform of claim 1 , wherein the control workflow comprises at least one of a state-based control workflow, an extremum seeking control (ESC) workflow, a proportional-integral (PI) control workflow, a proportional-integral-derivative (PID) control workflow, or a model predictive control (MPC) workflow to transform the timeseries of data values into the control signal using a control technique.

7 . The web platform of claim 1 , wherein the instructions cause the one or more processors to select the control workflow by:

analyzing inputs of the plurality of control workflows including the control workflow; and

selecting the control workflow from the plurality of control workflows in response to a determination that the input of the control workflow utilizes the timeseries of data values.

8 . The web platform of claim 1 , wherein the control signal includes a timeseries comprising a set of control signal samples.

9 . The web platform of claim 8 , wherein execution of the control workflow comprises:

transforming one or more samples of the timeseries of data values into one or more samples of the control signal by applying the one or more samples of the timeseries of data values to the control workflow; and

assembling the one or more samples of the control signal to form a control signal timeseries.

10 . The web platform of claim 1 , wherein the instructions cause the one or more processors to execute a proportional-integral-derivative (PID) control workflow to:

generate an error timeseries comprising a plurality of error samples, the plurality of error samples indicating a difference between one sample of the timeseries of data values and a corresponding setpoint; and

generate the control signal by applying a set of PID control operations to the error timeseries.

11 . The web platform of claim 10 , wherein applying the set of PID control operations to the error timeseries comprises:

generating an integrated error timeseries based on the plurality of error samples of the error timeseries;

generating a derivative error timeseries based on a change in value between consecutive samples of the error timeseries;

calculating a proportional gain component by multiplying the error timeseries by a proportional gain parameter;

calculating an integral gain component by multiplying the integrated error timeseries by an integral gain parameter;

calculating a derivative gain component by multiplying the derivative error timeseries by a derivative gain parameter; and

combining the proportional gain component, the integral gain component, and the derivative gain component to generate the control signal.

12 . A method, comprising:

receiving, by a web platform comprising processors coupled with memory, a timeseries of data values from a piece of equipment of an environment located remote from the web platform;

looking-up, using a timeseries identifier of the timeseries of data values, by the web platform, a control workflow from a database storing a plurality of control workflows for different pieces of equipment, the control workflow uses the timeseries of data values as an input and defines one or more processing operations to be applied to the timeseries of data values responsive to a detection that the control workflow executes for the piece of equipment;

executing, by the web platform, the control workflow based on the timeseries of data values to generate a control signal; and

providing, by the web platform, the control signal to the piece of equipment of the environment, the piece of equipment operating based on the control signal.

13 . The method of claim 12 , comprising:

storing, by the web platform, the database storing the plurality of control workflows comprising the control workflow on one or more storage media, the control workflow of the plurality of control workflows to execute for the piece of equipment and a second control workflow of the plurality of control workflows to execute for a second piece of equipment; and

selecting, by the web platform, from the plurality of control workflows, the control workflow that uses the timeseries of data values as the input.

14 . The method of claim 12 , wherein the timeseries of data values include a plurality of feedback values measured by a sensor of the piece of equipment of the environment.

15 . The method of claim 12 , comprising:

identifying, by the web platform, one or more other timeseries required as inputs to the control workflow, wherein the one or more other timeseries comprise a setpoint timeseries comprising a plurality of setpoint samples, the plurality of setpoint samples defining setpoints corresponding to samples of the timeseries of data values; and

generating, by the web platform, an enriched control workflow comprising the control workflow, the timeseries of data values, and the one or more other timeseries.

16 . The method of claim 12 , wherein the database is a directed acyclic graph (DAG) database storing a plurality of DAGs, each of the plurality of DAGs defining a particular control workflow;

the method comprising:

determining, by the web platform, whether any of the plurality of DAGs stored in the DAG database are configured to use the timeseries of data values as a DAG input.

17 . The method of claim 12 , wherein the control workflow comprises at least one of a state-based control workflow, an extremum seeking control (ESC) workflow, a proportional-integral (PI) control workflow, a proportional-integral-derivative (PID) control workflow, or a model predictive control (MPC) workflow to transform the timeseries of data values into the control signal using a control technique.

18 . The method of claim 12 , comprising:

analyzing, by the web platform, inputs of the plurality of control workflows including the control workflow; and

selecting, by the web platform, the control workflow from the plurality of control workflows in response to a determination that the input of the control workflow utilizes the timeseries of data values.

19 . One or more non-transitory storage media storing instructions thereon, that, when executed by one or more processors, cause the one or more processors to perform operations, comprising:

receiving a timeseries of data values from a piece of equipment of an environment located remote from the one or more processors;

looking-up, using a timeseries identifier of the timeseries of data values, a control workflow from a database storing a plurality of control workflows for different pieces of equipment, the control workflow uses the timeseries of data values as an input and defines one or more processing operations to be applied to the timeseries of data values responsive to a detection that the control workflow executes for the piece of equipment;

executing the control workflow based on the timeseries of data values to generate a control signal; and

providing the control signal to the piece of equipment of the environment, the piece of equipment operating based on the control signal.

20 . The one or more non-transitory storage media of claim 19 , where the operations further comprise:

identifying one or more other timeseries required as inputs to the control workflow, wherein the one or more other timeseries comprise a setpoint timeseries comprising a plurality of setpoint samples, the plurality of setpoint samples defining setpoints corresponding to samples of the timeseries of data values; and

generating an enriched control workflow comprising the control workflow, the timeseries of data values, and the one or more other timeseries.