Using Wireless Technology to Create Simple Laboratory Tools

Editor's Notes

• #3: We’ve seen in recent years that mobile technology has become more visible in the higher education sector. This is particularly the case in teaching through tools like moodle – blackboard, lecture videos and other online collaborative tool. However, the uptake of new technology seems to be somewhat lagging when it comes to research. The tools we use today are often the same as those we used a decade ago. The use of mobile technologies often present a cost-effective and portable alternative to existing tools that also potentially enable a greater efficiency in our research practice. And today we are going to present a case study that has investigated the use of a tablet based process control platform that connects to wireless sensors.
• #4: And just briefly – the process control system that we’ve used has been supplied by a company called Instrument Works. They are a sydney based startup company They have developing a range of wireless sensors that are able to connect To their data collection and process control software platform The company was created by a number of researchers who left academia to start the company With an aim to build better tools for engineers and researchers to collect and manage data That embraces new technology particularly the use of mobile devices.
• #5: So in term of context and motivation for this work – an looking at how mobile technology might help us improve our efficiency in our research and focusing on the field of engineering. Its important to look at the problem we are trying to address. Research – whether it be academic or commercial is a slow process Its takes a long time to go from having an idea – to having the results to validate (or invalidate) that idea We start with planning out ourexperiments / doing the experimental design. We then go an obtain all of the parts, tools, sensors and equipment. One interesting fact here is that in academia 80% of researchers report that they’ve had to beg/borrow and steal for equipment to complete there research. We then assemble our apparatus Setup our data collection and where relevant our control processes And its only then that we can conduct our experiments and start to obtain results And this isn’t a straight forward linear processs. We often have to iterate to get everything right, for the right results. And it can take upwards of 6 months to get this process right. What we’ve also seen in the past decade or so it that the availability and cost of sensors has reduced significantly meaning we are now able to measure more parameters than previously would have. This invariably makes experimental setups more complex and is generating signicantly larger amounts of data for us to analyse. Researchers aren’t able to manage all of these resources with the tools that we have available And this not only affect the quantity of research we can complete – but also the quality of that research. And by way of examples that you commonly see.
• #6: If you walk in to any research lab in any engineering faculty any where in the world what you see is this: Researchers using tools that were developed 10-15 years ago. The ability to collect data from these types of devices in limited Such that the most common use case is for the data to be collected by hand This typically results in two problems Transcription errors either when writing the data down, or when subsequently transferrring them to a digital format. The lack of collection of other metadata which can be used to ensure the results are valid For example in this picture we have a research recording data from a pH meter. Whilst the values are recorded what hasn’t been recorded are the Details of the calibration, when it was calibrated, details of the device or the probe or the calibration solutions Or perhaps she hasn’t recorded the temperature of the solution And what this means is, if we later discover that there was a problem with the probe or the calibration – we aren’t able to go back and identify the data that may be contaminated Or if we see some anomalies in the data aren’t able to look at some of those external influences that may have created that problem
• #7: Or at the other end of the spectrum – you might see this. An elaborate experimental apparatus – That had a lot of money spent on it to be built thats been designed for a specific project / experiment or purpose but That, isn’t flexible as if and when your experiments change Or the experimental parameter or conditions change. And significantly has been constructed in such a way that the users/researcher aren’t necessarily familiar with and are able to change/update or Or the conditions change that you want to run those experiments That uses tools that researchers aren’t familiar with and can’t modify What you also find with these types of setups – Is that when they are finished the project that they were designed for or the person that used them leaves They end up sitting in the corner of the lab gathering dust Or have some of their components removed and repurposed for other project – essentially rendering them useless for the next project that comes along.
• #8: And so based on this – we’ve been and dicussed these observations and conducted a survey with a number of engineering researchers who are predominantly PhD or Honours students. What we see is that for PhD Students about 90% percent of them either build their own equipment or adapt what they are using from existing equipment. And this makes sense when you consider much of what they are doing is extending the work of those before them. But for honours students we see that very few build there own equipment – and this might be considered due to the time contraints of an honours project. Most adapt it from other existing equipment similar to a PhD student. But a significant portion outsource their equipment building. And when you drill in to this you find that the outsourcing is predominantly to the PhD student or post-doc that they are working most closely with.
• #9: We also looked at the software that they use to collect data and control their equipment. For the PhD students – LabView is the tool that is most commonly used, follow by proprietary software rom the hardware suppliers, with a few using matlab – an in many cases they reported needing to use mulitple tools for collecting data. For example – some sensors in a experiment being connected to labview and others using proprietary software. But what also stands out is that in almost all cases they are using some tool to help them collect their data. On the other hand for the Honours students half of them, don’t use any tools and the other use either proprietary software or labview. So what we see is that for an honours students are more likely to manually collect their data – hand writing out their data. So what are the reasons for this difference
• #10: The first difference is that honours students only have a short period of time in which to do their research, typically only 10-14 weeks, compared to a PhD student which has 3 years. And in that time they need to design, setup, run and analyse their data. So their projects are simpler The experiments are simpler They collect less data and they are shorter And this make it easier for them to collect data by hand rather than have to learn new tools For honours students this is often their first introduction to research, so their skills are limited to begin with. As a result the experiments tend to be simpler and shorter, which make it more convenient to collect stand
• #11: So an alternative approach – with the instrument works system is to use a device that our students already have, their smart phone – or provie a deidcated device such as an ipod or ipad. What that provides is a user interface that they are already familiar with. What also interesting with with mobile devices you notice is that most apps don’t come with manuals or instructions but a common design paradigm that makes them easy to use and require little or no training Which is important for students who don’t have the time to learn a new set of tools. And with wireless connectivity – in this case Bluetooth low energy – we are able to connect to a range of different sensors and collect all of their data in one place.
• #12: As a first case study – we’ve had a student trialing the system during their Honours project in which they were investigating the recovery of solvents from wastewater in petroleum processing. This was a particularly useful case for the trial as the experiment setup needed to be place in a fume cupboard due to the volatle components being treated, which made a wired DAQ systems that we might have otherwise used problematic.
• #13: This is experimental setup used in this project. Where we are measuring the suction pressure on the vacuum side of the membrane. We monitored the temperature of both the feed and permeate sides to ensure they remained at their set points And a balance was employed to measure the membrane flux. (Although this was laterremoved as we weren’t able to get accurate results due to the low fluxes and the air flow in the fume cupboard)
• #14: Once the apparatus was setup – putting in the required sensors and setting up the data acquisition system on the phone took less than an hour. Data was able to be recorded at set intervals of the students choice and alarms were used to watch the set points for the temperautre. At the end of each experiment data was able to be exported via email as a CSV file where it was imported in to Excel for data analysis.
• #15: What we’ve found so far from this case study and using this tool – and we should point out that this case study is on going It’s a shorter learning curve than existing tools such as labview, which we typically don’t or aren’t able to use with an honours project because they aren’t familiar with it. By digitising the data collection process we reduce the chance of transcription errors due to manual data collection We’ve also reduced the setup complexity through the use of wireless sensors – which makes the setup process quicker and reduces possible points of failure and the need for trouble shooting. And finally we’ve provided a unified collection tool for sensor data for this project. Where previsouly we might have recorded the pressure data online but the balance data manually.