DataFil - A Complete Signal Processing Toolkit for iOS

A convenient and reliable set of signal processing tools for developers to integrate into their project to better utilise the sensors on an iPhone.

Creating a filter that has the right characteristics can be an extremely time consuming task, and often requires knowledge of signal processing.

This library provides an alternative to that, with ready to use filters that can be customised and integrated into your application quickly and easily.

Key features include:

• A consistent, easy to use interface and simple data model
• “Drag and drop” functionality. Each filter can operate as a standalone class. No project imports or external dependencies required.
• The latest Swift 3 syntax.
• Concurrent calculations where necessary.

Basic Code Usage

There are two options here:

• Take the filter algorithm and use it as a standalone object with your existing project code. Best for users with signal processing knowledge or projects with their own sensor data handling set up.

• Use the data source and filter manager classes to completely handle sensor data processing, allowing raw and processed data to be accessed with a simple notification observer. - Best for new users/projects.

Option 1: Standalone Filter

In addition to the chosen filter add the following files to your project:

• accelPoint
• Enumerations

Here we will use the high pass filter.

Initalise the filter:

let highPass = highPass()

(Optional) Adust parameters as desired:

highPass.setParameter(parameterName: "SampleRate", parameterValue: 30.0)

Register a callback:

let update = {(data: [accelPoint])->Void in {
   for dataItem in data{
      print("Filtered acceleration in X is: \(dataItem.xAccel)")
    }
}   
highPass.addObserver(update: update)

Finally, to pass data to the filter to process:

let exampleDataPoint = accelPoint()
exampleDataPoint.xAccel = 10.0
highPass.addDataPoint(dataPoint: exampleDataPoint)

The filter will now execute the callback registered earlier when it has completed processing the passed in data.

Option 2: Data Source and Filter Manager

Initalise the data source manager:

let dcm = dataSourceManager(sourceId: "John's iPhone")

Disable uneeded data sources, start up the rest:

dcm.modifyDataSources(accel:true,gyro:false,mag:false)
dcm.initaliseDatasources()

Start filters:

FilterManager.sharedInstance.addFilter(name: Algorithm.HighPass)
FilterManager.sharedInstance.addFilter(name: Algorithm.BoundedAverage)

In the class to receive data, register observers for notifications:

For raw data:

NotificationCenter.default.addObserver(self, 
    selector: #selector(myClass.newRawData), 
    name: Notification.Name("newRawData"), 
    object: nil)

For processed data:

NotificationCenter.default.addObserver(self, 
    selector: #selector(myClass.newProcessedData),
    name: Notification.Name("newProcessedData"), 
    object: nil)

Where myClass.newProcessedData and myClass.newRawData are Objective-C functions that will be called with when a notification of new data is posted. They must accept a single notification argument. For example:

@objc func newRawData(notification: NSNotification){
    let incomingData = notification.userInfo as! Dictionary<String,accelPoint>
    let accelData = incomingData["data"]
}

Apple Watch Data Streaming

The toolkit also provides the complete set of code requried to stream sensor data from an Apple Watch to an Iphone. The filters are all compatible with the Apple Watch, and all of the code in the GlobalModel group is designed to run on either platform.

Basic Usage

To stream data from the Apple Watch to a paried iPhone:

Both devices must have a copy of the GlobalModel group available to them. Ensure this by checking the Targets of the group in the attributes inspector.

On the Apple Watch:

Start the data sources:

let dcm = dataSourceManager(sourceId: "John's Apple Watch")

Disable uneeded data sources, start up the rest:

dcm.modifyDataSources(accel:true,gyro:false,mag:false)
dcm.initaliseDatasources()

Begin pushing sensor data to the iPhone:

RemoteDataInterface.sharedInstance.publishOutgoingData()

On the iPhone:

Start listening for incoming data:

 remoteCommunicator.sharedInstance.start(deviceId: "device")

In the class to receive the remote data:

 NotificationCenter.default.addObserver(self, selector: #selector(myClass.newRemoteData), name: Notification.Name("newRemoteData"), object: nil)

Where myClass.newRemoteData is an Objective-C function that will be called with when a notification of new data is posted. It must accept a single notification argument.

Other Features

The Apple Watch communication shown here is extremely powerful, and allows for to way sending of key value messages between the devices, as well as the observation of incoming keyed messages from anywhere in the application. For more informaiton, see the Remote Communicator page in the wiki.

Showcase App and Example Code

Included in this project is a showcase app which can be used to experiment with the filters to find the right one for your needs. It also serves as example code of all the features detailed above.

Instalation

Requires Xcode 8.3 or later and a ** real ** iPhone or iPad running iOS 10 or later. Simply open the DataFil project in Xcode, sign it and sideload it onto your device.

Features

• Realtime sensor data display
• Interactive filters
• Data recording and exporting as CSV or JSON
• Apple Watch sensor data streaming for visualisation or recording.

A full user guide on all of the features of the showcase app can be found in the How to Use the Showcase App and

Using an Apple Watch with the Showcase App pages of the Wiki.

Licence

The filter code and showcase app project code is reelased under the MIT licence with the following notice:

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

Also, please note:

The implementation below is NOT designed to be used in any situation where the failure of the algorithms code on which they rely or mathematical assumptions made therin could lead to the harm of the user or others, property or the environment. It is NOT designed to prevent silent failures or fail safe.

This README, the Wiki and any other documentation is released under CC-BY 4.0