PowerSensr: the smart voltage logger

Back in 2014, I worked on one idea that had been in my mind since my first graduate job at powerPerfector.

Problem

powerPerfector was (before its pre-pack administration) a voltage optimisation technology company that helped buildings to become more energy efficient. By optimising the voltage supply to a building, you can minimise the amount of energy wasted by the building’s electrical devices.

One of the main stages in the sales process was the collection of voltage data at the site. powerPerfector used to post the client a voltage recorder such as a Fluke VR1710* data logger (“logger”). However, this process was very slow as it would take a few days for the logger to arrive, a week to collect the data, then a few days to post the logger back, before downloading the data back at pP HQ.

This was a best case scenario: frequently loggers would go missing at client sites, or they would be forgotten about or even sent back with no data on them. In addition, they required quite a lot of up-front investment to procure, meaning that the number of loggers powerPerfector could send out was limited and therefore delays were even greater.

Solution

I wanted to build a cheaper data logger with the ability to upload the data recorded directly into the cloud. This would allow much faster data collection and due diligence of sites, but also they could even be left in situ to help building managers ensure that their sites are compliant with power quality regulations.

Finally, leaving a logger in situ that can broadcast the current voltage level (of the lowest voltage part of the site) could have even been used in tandem with a dynamic powerPerfector (previously known as a pP+). The logger could have been used to provide a continuous signal to the control unit of the pP+ so that the voltage level would be best optimised for the whole building.

Product concept 

The working title for my product concept was “PowerSensr”.

I essentially wanted to build a single phase voltage and power quality logger similar to the Fluke VR1710* that had the ability to upload data into the cloud in near real-time.

I spoke with an electrical engineer to scope out the most appropriate technologies for what I wanted to do and to get a rough cost per unit. We came up with the following product specs:

  • WiFi for connectivity (3G to be investigated during market research but unlikely to be viable due to costs).
  • Amazon Web Services for the cloud platform.
  • Low internal memory requirements.
  • USB connection to be explored for backup purposes.
  • Low price point (£25-£125) to compete against £650 for the Fluke VR1710*.
  • Simple dashboard at first.
  • 1 second resolution on the measurements.

We discussed building a prototype, but the guy was too busy to help. However, he suggested that I could build the initial prototype based on Raspberry Pi with some guidance from him. Once the concept was tested I could contract the design from organisations such as Cambridge Consultants if I raised enough funding to pay their fees.

Sensr Initial Conceptual Design

PowerSensr Conceptual Design by Jack Raison, 2014

I was very lucky in getting the initial form concept designed by the award-winning graduate design engineer, Jack Raison.

Jack took the high-level design brief to produce the render of the device in the inset picture (left).

Reviewing the idea

I then used this information to review the suitability of the idea using the New Business Road Map.

First, I scanned the marketplace to assess the competitive landscape. The main competitor was Fluke with their Fluke VR1710*. It has many features (except cloud-upload functionality), is very robust but is quite pricey at around £650 at the time for bulk orders. Also, Fluke have a dominant market position as they are a well-known brand in the sector.

Having spoken to several distributors and resellers of this sort of equipment, I realised that buyers have a major focus on brand and quality for reliability. Building this trust for PowerSensr would be one major obstacle if I decided to move forward. Additionally, Fluke were in the process of launching Fluke Connect to bring cloud connection and monitoring features to some devices. The VR1710 was not one of them at the time, however.

I also spoke to 38 Zeros about their CloudLogger device. This was very interesting, as they operate the devices over the phone network and charge $199 for the devices with a 1 year contract of $15 per month (presumably the “Nespresso” business model of making a loss on the hardware to lower the sticker shock of the purchase price).

CloudLogger is focussed on energy consumption for Building Managers rather than voltage or power quality. Intriguingly, they only emailed the data to the customer rather than providing a dashboard solution as customers often have significant dashboard facilities already for their sites.

Once satisfied that there was nothing on the market that could provide what I wanted, I tested the concept for my product with target customers and organisations that would potentially partner with me to sell it.

The first ones were the distributions and sales organisations for this sort of kit. I was shocked to hear that they would only ever sell about 1 or 2 a month as this was quite a niche bit of kit. He also stated that people only really trust one or two brands in the marketplace, as building services engineers are very particular about their bit of kit.

I also spoke to a Director at a built environment consultancy who I knew from a previous role. He was surprised that I was spending any time on a product in this area, as he thought that “nobody really cares about power quality of voltage”.

Indeed, Ilian Iliev at EcoMachines Ventures asked me a great question during a conversation at one of his Office Hours that really stuck with me: “What is the value of the “cake” here and how much of a slice will you charge?“.

It occurred to me that, although voltage problems will only get worse as RE and switch-mode IT loads get added to the grid, the reality is that most people don’t really care about measuring this too much. Therefore, the size of the “cake” (i.e. the value provided to customers by this solution) was small.

The only real customers I had in mind for this technology were powerPerfector and the rival Voltage Optimisation/Management companies. However, pP went into a pre-pack administration in 201X and EMS Powerstar went through tough times after a massive collapse in the market.

Additionally, the risk that Fluke’s Fluke Connect project could encroach on this space seemed like a major risk even if I could build a device that could work at the accepted price point. Their brand awareness would blow me out the water.

I therefore came to the conclusion that this was not a big enough market to chase. In fact, I would even go so far as to state that I had a solution and was looking for a problem. So in late 2014 I decided to bin the project.