Sometimes you need to show a customer or your boss what's possible before you build it. We've created Synth as a powerful IoT device simulator to bring your use case to life before you've even deployed your IoT devices. We find this very effective for closing sales and now we're ready to share our secret so you can too.

How does it work?

Synth is an application which runs on any computer or in the cloud. It comes with a set of basic device behaviours which you compose together in a scenario to create simulated devices to show off your use case. Then just ask it to simulate a single device or a million, for either the past month or in real-time - and hey presto. 

Let's look at some of the demo use cases Synth comes with out-of-the-box:

1. Building Management Systems (BMS) 

The BMS scenario shows a single building containing many devices of different types including:

• Smoke detectors, which suffer from various faults including unreliable wireless communications
• Light bulbs which get switched on and off
• PV solar panels which generate electricity¹
• A weather sensor

Reading the dashboard from top left, we see:

• The location of every sensor on a floorplan (can show the live state of each sensor or its historical performance).
  
  
• The distribution of the received signal strength indicator (RSSI) for the smoke alarms - showing a classic distribution with most alarms, being in the middle of the range.
• The next chart along shows uptime against RSSI - the kind of analysis that DevicePilot makes easy but is impossible in other tools. Uptime (Y axist) is measured by whether we reliability hear the regular "heartbeat" signals from the smoke detectors, and this is then correlated with signal strength (X axis) to answer the question "does signal strength affect uptime?".
  
  The answer is yes: uptime is definitely affected as the strength falls below -77dBm. Perhaps not surprising as a general observator, but now we have the empirical, quantified facts based on devices in the field, so can calculate cost:benefit of any remediation. 
  
  
• The smoke alarms also suffer from a variety of error conditions, ranging from dirty sensors to being opened by unauthorized persons. Top right, we see a stacked chart showing day-by-day which errors are most common (making any sudden changes or trends visible).
  
  
• Bottom left, we see a chart of how many light bulbs are switched on, by time-of-day. This view, where, say, 30 days of data is wrapped around into a "typical day" is very useful for seeing daily patterns.
  
  
• We see a similar view for how much power the bulbs are consuming.
  
  
• And again, a similar view for how much power net is being generated or consumed by the building .

2. Mobile pallets

In this scenario, we're tracking a number of mobile devices around the UK. These devices are reusable pallets which tend to travel between a small number of depots and can build up in the same depots. 

• Top left, we see their current positions (clustered if many are in one place).
  
  
• Bottom left, we see only the pallets which have been stationary for more than one week. One of DevicePilot's most useful analytical powers is the ability to reason about time in this way, which - ironically - many time-series databases cannot.
  
  
• Top middle shows the number of pallets moving at any one time, day-by-day. Clearly, we're not making very good use of our assets if only about 2% are in use at any one time.
  
  
•  Top right shows how much time in total pallets spend in any one location (i.e. in which location are they building up? should we send a truck to recover them or have a chat with the local supervisor?). The reason that this stacked chart doesn't quite add up to 100% is those 2% or so which are in motion at any one time.
  
  
• The pallets also have shock and temperature sensors on them to detect if goods are being damaged in transit, so bottom middle shows the number of shock events per day.
  
  
• Bottom right shows how much time pallets are spending in an "over temperature" situation - potentially causing damage.

3. Smart meters 

These smart meters are scattered around India (interactive map, top left).

• In general, they have good uptime and we can see their uptime groups by vendor (top middle), but clearly one vendor is having problems.
  
  
• Top right, we can see a time-based view of this, and there there seems to be some recurring problem with that vendor.
  
  
• Bottom left, we see a stacked fault chart (like the smoke alarms above) and bottom middle shows faults by geographical region.
  
  
• Bottom right pulls out the specific devices with faults which belong to the problematic manufacturer.

All DevicePilot dashboards are live and interactive, so these pictures will change minute-by-minute to reflect changing reality (well, simulation in this case).

4. Sump pumps 

This scenario was built to show how DevicePilot can be configured to produce interactive dashboards from a standing start in just 15 minutes.

The use case here is that roads have "sumps" next to them - big trenches that rainwater can quickly drain into, and which then needs to be slowly pumped away.

• Top left, we see that there are 10 sumps in Cardiff, Wales and 10 sumps in the Netherlands. 
  
  
• Top middle shows how much rainfall is being detected in each location (and in this simulation this is driven by real rainfall in these locations).
  
  
• Top right shows the amount of energy (and therefore money) the pumps are consuming to clear the sumps. Note how the blue bar levels off at the right hand side - this is because the pumps in Cardiff are running flat out and still not clearing the sumps. 
  
  
• Bottom left shows us the maximum levels that the sumps have reached in recent days, and bottom middle highlights how much time the pumps are running flat out. Bottom right then highlights how often the sumps have been overflowing (i.e. the sump level is at a maximum) which tallies with those pumps running flat out.

5. Vending machines 

This scenario shows a set of vending machines across the USA².

• Top left, we see a heatmap showing density of machines.
  
  
• Top middle, we see how many vends per day (i.e. revenue-generating activity) and how much replenishes a day (i.e. cost-driving activity).
  
  
• On the right hand side, we see a list of hte machines with the highest level of cash in their cash boxes (a potential business risk).
  
  
• Bottom middle, we see those machines which have low availability of goods to vend (damaging revenue) and bottom right, we see a list of the specific offending machines with low stock levels, sorted so the worst ones are at the top. This is then an "action list" of problems to solve.

6. Drug fridges

This scenario simulates temperature and presence sensors from our partner, Disruptive Technologies. We see a number of fridges containing drugs placed around a hospital.If the fridge doors get left open, the fridge warms up and damages the drugs within - meaning the potentially life-saving drugs won't work, but make you worse.

• Top left, we see the bad news - how many "drug ruining" events have happened recently (defined as the fridge warming up beyond a certain temperature). Below that is a time chart multiplying each event by the cost of the fridge contents ruined. Bummer.
  
  
• The next column across shows the maximum temperature of any fridge over the last fortnight, and the number of fridges with a temperature anomaly (above normal operating temperature) and the number of fridges with a temperature problem (temperature so high it will damage the drugs). At the bottom of this column, we see which specific devices have had temperature anomalies. 
  
  
• The middle column shows a floor plan, with colour representing the live temperature in each fridge (and the exact number appears as a tool tip in the software if you hover over it). Below that, we see a measure of how many "door events" (door openings) there are per day - presumably correlated with levels of use of the fridge. And below that, we see the pattern of the typical day - mostly the doors are opened in normal office hours. 
  
  
• The column on the right analyzes how much - and for how long - doors have been left open. Number of fridges with doors left open for two minutes, five minutes, and 15 minutes, and bottom right the specific ridges which have had doors open for more than 15 minutes (time to pay that clinic a visit!). 

Ready to get your hands on Synth?

Simulated data can bring IoT solutions to life, accelerating project go-ahead and closing sales.

You can get your own copy of Synth - it's completely free and open-source.

Full documentation here

Full code is here

Coupled with DevicePilot, you can get your own use case up and running in hours, showing it off to your customer or boss with deep analytics and interactive visualization. Enjoy!

^{1. Synth realistically simulates light levels and weather based on location and time}

^{2. Synth can scatter devices randomly but realistically across a chosen area, using population density as its guide.} 

Further reading: 

Beautiful interactive dashboards in just 15 minutes

How to get SCADA working with IoT

4 architectural choices you'll need to consider