Intelligence at the Analytical “Edge”

This is my first ever blog post, so a little grace in advance, please.  I’ve been thinking lately about parallels between historical developments in the networking and computing fields and current developments in the analytical chemistry field.

Most of us are familiar with the sea changes from mainframes to minicomputers to desktop PC’s to smart phones, with the result that an iPhone 5S has about 1000 times the compute power of the first Cray 1 supercomputer.  Similarly, networking intelligence in the form of TCP/IP routers moved from the Arpanet core to enterprise networks to your home DSL or cable modem.    In each case, the move of intelligence to “the edge” has been enabled by robustness, simplicity, standards, system architecture, processing power, and software.

The analogous analytical chemistry “network” is the hierarchy of labs from central R&D at the top of the pyramid, to method development, to QA, to at-line, to on-line.  (I think some in the industry refer to movement between these categories as “functional migration.”)  Like the TCP/IP gateways and supercomputers in the early Arpanet core, today’s R&D and method development labs have a very rich set of technologies available.  These include sophisticated and expensive instruments like triple quad MS, GCxGC-MS, TOFMS, Orbitrap, etc.  Many of our early VGA-100 detector installations have also been in these high-end R&D and method development facilities, to complement very sophisticated capabilities already in place.

The challenge with moving most of these technologies down the pyramid of lab types – or to “the edge” in the networking analogy – is lack of some of the aforementioned attributes: robustness, simplicity, standards (that can be implemented successfully by available personnel), and system architecture.  Processing power is available, but not the limiting factor, and software exists, but perhaps not in the right paradigm.   VUV technology, on the other hand, has already started to get traction in QA and production applications, in addition to R&D and method development labs, because of its inherent attributes.

Simplicity

VUV absorption spectroscopy is similar in complexity at the user level to UV-Vis absorption spectroscopy.  The complexities of dealing with VUV light, the algorithms, etc. are all engineered up front by VUV Analytics.  Unlike sophisticated mass spec analysis, the instrument finds a unique spectral signature for each compound, so there is no room for undiagnosed user error, bad assumptions that skew results, etc.

Robustness

There is no vacuum pump, few moving parts, and no consumables beyond familiar items like columns, lamps, and inlet liners.  There are no valve timings or traps to calibrate.  The instrument self-calibrates to light, dark, and the carrier gas on every run.

Standards

The first of many ASTM standards is in draft form in a working group now.  Our ongoing approach, as with this gasoline PIONA+ standard, will be to show equivalency to existing ASTM and other standards.  Stay tuned for new application announcements early in 2016.  Importantly, standards based on VUV technology won’t need an R&D guru to run them to get correct results.

Systems Architecture

Here’s where the analogy to networks gets more interesting.  We think an R&D or method development lab with a variety of sophisticated hyphenated techniques, like GCxGC-TOFMS, can use those capabilities alongside GC-VUV to develop simple and highly automated GC-VUV methods that can be propagated to the analytical “edge” for routine analysis.  Our ASTM draft method and PIONA+ software in beta is a concrete example of this, with many more to come.

Processing Power and Software

The VGA-100 produces a very rich, unique, multi-dimensional data set with each run.  Our challenge is to make it as easy as possible to enable users to draw conclusions from the data and/or automate the analysis.  Although our company spends as much effort on software as anything else we do, we have only begun to scratch the surface of what we can do, both in real time at the instrument and offline post-processing.   Don’t be surprised if our next product announcement is software, and it certainly won’t be the last one.

Okay, enough infomercial on VUV technology.  What are the key takeaways for analytical chemistry leaders?

• Like the networking field, a systems of labs at a big organization has an architecture that is hierarchical.
• Over time, the most significant performance improvements come from pushing intelligence and processing power to “the edge.”
• When investing in sophisticated new capabilities for R&D and method development, attention needs to be paid to how easily they can be propagated to the analytical “edge”, where simplicity, robustness, and user-friendly standards are essential ingredients.

Click here to download the VGA-100 brochure for more information on detector specifications and analytical characteristics.