What's in your lab?
March 27, 2018 9:10 AM   Subscribe

For something I'm writing, I'm looking for examples of the complexity (and expense, if you have any details to share) of the type of lab equipment used to research and develop complicated new technology that addresses major world challenges. The tech might be radically more efficient solar cells, a new treatment for Alzheimers or cancer, nuclear fusion, the hyperloop, etc.—anything with lofty aims. Can you share an example of what's needed to develop a particular technology? I want to contrast developing this type of technology and developing something like an app that requires very little equipment at all.
posted by pinochiette to Science & Nature (13 answers total) 9 users marked this as a favorite
Best answer: I work on guidance systems for air and space vehicles (so "lofty" but maybe not your definition of the word). We have several different labs:

1)Hardware-in-the-loop integration lab. The guidance system is split apart into individual electronics cards, so we have fixturing for all those parts. Power supplies/power conditioners, data logging tools, GPS emulators, "live sky" receivers, vehicle system simulators, cooling systems, banks and banks of computers for various tasks.

2)System integration labs. The guidance system is tested as a whole unit, so we have accurately surveyed "piers" to mount them to in addition to the usual power supplies, cooling supplies, emulators, and user interfaces. There are some other troubleshooting labs that have very specialized equipment for testing the system.

3)Environmental test facilities. The equipment has to go on vehicles that go through severe environments, so in addition to all the above equipment that makes the system work, we also have shaker tables, temperature chambers, vacuum chambers, and centrifuges (including the largest centrifuge on the East Coast). We can simulate the air pressure and vibration and shock environment, combined, for a full rocket launch mission.

At a subsystem level, we have things like clean rooms, microelectronics labs, and specialized machine shops. We have some mobile "labs" as well - ground vehicles, a pod that we mount to an F-15, and even a test cell on a boat. Every lab that we have (well, except for the mobile ones I guess) require precise temperature and humidity control, as well as ESD (electrostatic discharge) and sometimes FOD (foreign object debris) protection.

Costs I'm not too familiar with. Simple shaker tables of the size that we use can run around $500,000 new, and we have some very sophisticated shakers. Ballpark for all of the capital test equipment we have is probably in the hundreds of millions of dollars if it all had to be replaced today.
posted by backseatpilot at 9:28 AM on March 27, 2018

Best answer: I do research on human visual perception, focused on driver behavior in manually-controlled and autonomous vehicles. The capital equipment for this kind of work can be things like full-cab driving simulators (think single-digit millions to set a good one up from scratch, more if you want more immersion) to instrumented vehicles (depending on what you want to do to it, anywhere from 50k to 500k+) to autonomous testbed vehicles (maybe 100k at the low end, but probably several hundred thousand at the higher end).

If you wanted to set up a good driver behavior / autonomous vehicles lab, you're probably looking at tens of millions in both equipment and staff (I've watched one automaker do it recently, and it's a lot of people and a lot of money).
posted by Making You Bored For Science at 9:39 AM on March 27, 2018

Best answer: I worked in a molecular biology lab (and still work adjacent to one). Here's the kind of equipment I used or worked/work with people using on a regular basis:

- Fume hoods, biosafety cabinets, and cold and warm rooms allow working under different temperature and sterility conditions.
- Incubators, incubator/shakers, and freezers at different temperatures (4C fridge, -20 freezer, -80 freezer) allow storage of samples at different temperatures for cell growth, chemical reactions, or safekeeping of samples.
- PCR cyclers, which run temperature programs to make copies of DNA
- DNA sequencing equipment of a variety of levels of complexity/cost
- Fluorescent microscopes of, again, varying complexity/cost
- Mass spectrometry equipment for measuring proteins
- Glassware, disposable plasticware, and the equipment to clean & sterilize it
- A nice computer cluster to crunch the data (although some of that is done on the cloud now, sometimes with donated compute time from the companies involved)
posted by quaking fajita at 9:41 AM on March 27, 2018 [2 favorites]

Best answer: I've worked in a variety of labs aiming to either improve our understanding of hearing loss or how we treat it (hearing loss is the third most common chronic condition in the world and we aren't all that great at fixing it).

-The most expensive pieces of "equipment" are the spaces. Our anechoic chamber is a highly specialized large room built underground, designed to have no reverberation. Add in a bunch of super expensive speakers and you're looking at millions and millions of dollars. Above ground sound booths are also very expensive to purchase and install.

-Equipment used to test hearing and hearing technologies is crazy expensive. Audiometers, brain stem response rigs, EEG equipment, otoacoustic emission recorders are all super expensive and all involve incredibly expensive and advanced microphones and earphones/speakers and computers. Even some of the cords we use are thousands of dollars a piece. Our soundcards can be 10-20k. Test boxes to measure hearing aid output are 10-15k.

-The calibration equipment which we just use to check our other equipment is really advanced and expensive. The kit can cost $15k.

-We have a driving simulator to test how people are hearing when they drive. No idea how much this costs but probably not cheap.

-We have specially designed mannequins which are designed to mimic the mechanical acoustics of the ear and can record signals with the same acoustic transforms as a live human.

-Couple this with hundreds of smart phones, LENAs (devices which will record an entire day's worth of conversation and then analyze it to tell you things like how many conversational turns were taken, etc), hundreds of hearing aids and implants, it's all quite a lot.
posted by Lutoslawski at 9:47 AM on March 27, 2018 [1 favorite]

Best answer: In a Genetics Lab:
When I was interning in a genetics lab in the 1990s, we had mice. Lots and lots of mice. Rooms filled with mice. Lab ordered, honest to goodness meeting your exact genomic specifications mice. That meant we could ask for a left handed mouse with hip dysplasia and a predisposition to obesity. The implication is that you had a metric ton of support staff, ready to take care of, feed, deliver, sex, impregnate and sort mice for you - which is hella weird.

The lab was sterile. It was minimal personality. Personality was in people's offices.

In an Engineering Lab:
Back in the early 2000s, we had a C&C milling machine and a full machine shop to make the parts we needed for prototype. These days, someone might use an industrial 3d printer to make the same thing.

We also had access to forklifts, tons of tools, and a tig welding station. (Wrenches, drivers, hardware, cutters, plasma cutters, scribing tools,)

One of my favorite things we had was a large metal vertical plate on locking wheels that was about 8' tall x 5' wide and 1" thick. This was for when projects went awry and launched out at unexpectedly high velocity. They'd smack the portable wall instead of killing someone 1000 yards away... Yes... it had a few *big* dents in it.

In an Electronics Lab:
We had a lot of partial prototypes, boards built but with an error in the silica that failed to meet specifications. We had an assortment of oscilloscopes, frequency sweepers, leads, common chips, soldering stations, and so on. These things have always been acquired over time - which means there are some *better* and preferred pieces of equipment that you either claim for yourself and set at your specific work area, or are general use, but are fought for on a first come first serve, first forgotten scenario. There is always a radio. Whoever controls the radio likely has better gear than they need, but they have the clout to hold onto it.

We had 3 or 4 lab projects - things that we worked on, that just wound up living in the lab and could be demoed as needed for clients that might be on a tour. When these weren't in use, sometimes they were great for table space.

A lot of floor space with old tape marks on it. Different projects required different layouts of the lab, so these helped us reconfigure as needed. Light tables for schematics, drawers for common parts, spools of wire in various gauges mounted on a wall. These were clean but messy places. Everything had a place, but since there was so much stuff that you might need, there was just so much of it that there was a lot of ... visible conflicting personal organic organization.
posted by Nanukthedog at 10:40 AM on March 27, 2018

Best answer: This article discusses start-up costs in academia and might give you a sense of scale. Note that startup costs are just what you need until you, presumably, write enough successful grants to maintain staff, equipment, etc. yourself, and does not include any shared university resources that the person may be hooking into. So I'd call those numbers an extreme low end to get started independently, and would figure that 25% of those costs per year to maintain would be on the low end to continue. For an independent researcher, this would be roughly the "one person in their garage" level of engagement.

I work in a shared university facility and we have about *mentally does inventory* 9 million dollars' worth of equipment, with maintenance costs in the 100's k per year. The costs are shared by many researchers, but I know a number of university research labs with probably 1-2 million dollars' worth of equipment, and 100-300k staffing and maintenance costs per year. The cost for infrastructure-related stuff, HR-type stuff, and consumable supplies are probably half that. A startup company would pay way more for all of those things (grad students are cheap; universities have huge buying power and get discounts.)

I've been in the room where the instrumentation I use most is manufactured. It takes 3 months to build one, and there is space for about 15-20 of them on the build floor. Some of the bits are made in-house, but lots of parts are purchased from elsewhere, even though some of those parts are so specialized that this company is the only one who buys this specific widget with these specific parameters (those places presumably have similar looking rooms). Most of the people involved with the manufacture have advanced degrees or very specialized skill sets. A bare-bones version of the machine costs about 1 million dollars.

On a larger scale, the Large Hadron Collider cost less than 5 billion to build and costs about 1 billion per year to operate.
posted by tchemgrrl at 10:54 AM on March 27, 2018 [1 favorite]

Best answer: Maintenance contracts on the equipment. Supplies and dyes and primers for the sequencing equipment, gels for PCR, lightboxes, pipettes and tubes. Gloves and glasses and gowns for the bench staff and visitors. Cleaning supplies. Pens and logbooks, printers and ink cartridges and paper (you'r never as paperless as you try to be).

And the modeling is for equipment only? Multi-person licenses and fees to access the databases needed to interpret what the equipment puts out are not being assessed?
posted by beaning at 11:45 AM on March 27, 2018

Best answer: I work in a lab that does a lot of molecular microbiology and vaccine research on bacterial pneumonia and respiratory infections. In addition to standard equipment like centrifuges, refrigerators, freezers, incubators we have two high-throughput DNA extraction machines (about 60K each), a real-time PCR machine (about 40K), a spectrophotometer (8K), gel box camera system (8K), nanodrop (5K) and a Microarray scanner (150K). These are all pretty basic instruments. Note that the stabdard lab stuff is also pretty pricey- a -80C freezer costs about 16K and we have three of them. The immunology lab next door also has a luminex and a FACS machine, which are common and pricey items (over 100K). All costs are approximate, lab in Australia.
posted by emd3737 at 11:48 AM on March 27, 2018

Best answer: A list I once made to outfit a lab that did protein biochemistry on proteins mainly related to cancer. This list was made to set up a new lab in 2005, so some of it is a bit out of date.
Laminar Flow Hoods
Low Quality Microscopes
Low speed centrifuge
Roller bottle incubators
storage for media - fridge/freezer combo
suction flask set up
Vortex mixers
water bath

37oC incubator
4oC fridge
bunsen burners
chilled microfuges
chromatography - needles
chromatography case
chromatography loops
chromatography set up
chromatography tubing and connectors
geiger counters
gel boxes
Gel camera system - digital
heat blocks
heated stir plate
high speed centrifuge
hybridization oven
ice buckets
Light box
millipore unit
minus 20oC freezer
minus 80oC freezer
multichannel pipetors
northern vacuum transfer apparatus
osmolarity meter
PAGE vertical gel boxes
pH meter
power supplies
radiation disposal boxes
radiation plexiglass shields
Realtime PCR
reepat pipetors
shaking incubator
sonicator - bioruptor
Spectrophotometer - vis and UV
step stool
stir bar
stir plates
transfer apparatus
UV box
vortex mixers
Water baths
white boards

UV cross-linker
fluorescent microscope
fax machine
xray film developer

acetic acid
boric acid
bromphenol blue
Citric Acid
glycerol phosphate
isoamyl alcohol
methylene blue
Naphosphate dibasic
Naphosphate monobasic
PEG set
ponceau red
Sodium azide
sulfuric acid
xylene cyanol
acrylamide:bis mix
Bradford reagent
phenol (diff pH)
Salmon sperm DNA

LB mix

colloidal blue
competent cells: XLI, DH5, BL21
dual luciferase
lipofectamine 2000
TA blunt cloning

15ml tube racks
50ml tube racks
aluminum foil
autoclave gloves
autoclave tape
flat tweezers
freezer boxes
freezer gloves
hybond N+
lab coats
microfuge tube racks
phosphorimager cassettes
plastic wrap
PVDF membrane
razor blades
tygon in various sizes
vacuum grease
xray film
xray film cassettes

10cm plates
10ul filter tips
10ul tips
14ml round bottom tubes
15cm plates
15ml falcons
1L filter units
1ml filter tips
1ml tips
200ul filter tips
200ul tips
20ul filter tips
20ul tips
24 wells
500ml filter units
50ml falcons
6 wells
chromatography tubes
PCR tubes
plastic spoons for weighing
QRTPCR plates
roller bottles
scintillation vials
sorvall 35ml tubes
UV cuvettes
weigh boats S/M/L
weigh paper

100ml grad cylinder
1L beaker
1L erlenmeyer
1L grad cylinder
200ml beaker
250ml grad cylinder
2l beaker
2L erlenmeyer
4L beaker
4L erlenmeyer
500ml beaker
500ml centrifuge bottles
500ml erlenmeyer
500ml grad cylinder
centrifuge tubes
cloning cylinders
glass petri dishes
hyb bottles
pasteur pipets
pyrex dishes

get starter kit?
Pfu turbo
protease inhibitors
Proteinase K
RNase A
RNase H
RNase inhibitors
Shrimp alkaline phosphatase
T4 DNA ligase
T4 polynucleotide kinase
T4 RNA ligase
Taq polymerase
Vent polymerase (or other Pfu kind)
Klenow exo-
Dnase I

FLAG peptide
FLAG resin
mouse secondary
rabbit secondary
GST resin
protein A resin
protein G resin
Nickle resin


2dary antibodies
coverslips - square and round
mounting media
slide cases
slide sleeves

allen wrench set
first aid kit
screw driver set - phillips and flat

blank CDs
hole punch
lab notebooks
lab tape
manilla folders
mechanical pencils
packing tape
pads of paper
post its
printer paper
scotch tape
sheet protectors
toner cartridges
white board erasers
white board markers

bottle brushes
dish soap
dishwasher soap
hand soap
radiation soap
scrubby sponges
trash cans
posted by sciencegeek at 2:37 PM on March 27, 2018 [1 favorite]

Best answer: So one thing about research scientists is they love gloating about their equipment and capabilities. See, e.g., this page from the National Renewable Energy Lab about their photovoltaic characterization capabilities. I'd ballpark that each of those experiments represents at least 200 sq ft of lab space and around a quarter million dollars of equipment (more likely 10-50x that).

Take a look at the "temperature range" column. One experiment lists that it goes down to 77K, that's the temperature of liquid nitrogen. A lab like this is going to have a big tank of liquid nitrogen somewhere in the parking lot. Liquid nitrogen itself is pretty cheap (cheaper than milk or gasoline) but the tank is expensive and there are evaporative losses. It's replenished daily or weekly by a big tanker truck from a gas supply company.

An experiment like that isn't going to consume enough nitrogen to justify a dedicated line from that tank, so there's probably a separate truck from the same company that comes with a bunch of smaller 160 liter dewars that can be hand-carter around to where needed. That same truck will also bring the other bottled gasses the lab uses: argon, hydrogen, helium, oxygen, CO2, probably halogens and silane and some others. A lot of these are used enough that they're plumbed into the building as well, along with compressed air and a few other things I'll get to in a minute. A facility like this will generally have at least one pipefitter on staff.

The other experiments go down to 4K, that's liquid helium temperature. You can get liquid helium delivered too, though it's incredibly expensive (ten years ago we paid about $1200 a bottle, it's probably more than that now). A lot of apparatus that requires those temperatures uses a closed-cycle system instead, which includes a component like this. You can see the size of the power cord coming off that, and the requirements listed below: 25A of three-phase. That works out to around 12 kW. Now, that's peak inrush, and not steady-state operating power, but still well more than the typical residential home ever uses. Hell, even a lot of light industrial real estate isn't wired for three-phase. So a facility like this will also have several electricians on staff and a surprising density of electrical receptacles on the walls and conduit running alongside the plumbing.

Oh, speaking of which, see those brass fittings to the lower left of the unit? Those are for cooling water, which is also plumbed in a building like this. It's not actually water, it's a glycol solution that's chilled in rooftop units and pumped through a pressurized closed circuit to where it's needed. Generally alongside the send and return lines for that you'll have purified water from a giant RO system somewhere, and compressed air for actuating valves, floating vibration-isolation systems, and general utility purposes like that.

So that's all well and good for cooling the sample to measurement temperature, but you're also going to need to thermally insulate it from its surroundings. You generally place it in some sort of vacuum chamber, like those offered here. That's a big stainless steel tank with flange fittings for attaching vacuum pumps, airlocks, viewports, manipulators, and various instrumentation and fluid and electrical feedthroughs. Those chambers have to be pretty thick to withstand atmospheric pressure and correspondingly weigh a lot. If you're in a room with one, it's pretty common to see an engine hoist or floor crane stowed against the wall; the nicer facilities generally have a gantry crane built into the ceiling. Obviously then you're not just going to set a system like that on the lab bench: you're going to to fabricate a custom support frame, so you're going to need a machine shop and a metalworker or two on staff as well.

For evacuating the chamber, you're also going to need a vacuum pump. Sometimes you can get away with a freestanding one placed next to the apparatus, but there are a couple problems. First, they generate a lot of vibration, which can interfere with measurements and processes. It's generally better to put it in a different room, preferably on a different foundation slab or otherwise structurally isolated. Second, liquid helium temperatures can cause a lot of ice to form out of hydrogen and other nasty things you don't want to be pumping into the room once it warms back up. So generally the room you put the pump in has a big vertical ventstack that inducts the pump exhaust above the roof of the building (these are visible in another picture from the NREL site, here).

That should give you an idea of the sort of facilities necessary for conducting a research effort in photovoltaic technology. You tend to find similar plant in other areas of semiconductor, nanotech, and materials research. The architectural requirements are such that the buildings are generally purpose-built (though retrofits are possible... if you're ever in Kendall Square by MIT and you look around you'll see lots of old building with exhaust stacks like those I described).

In semiconductor production economics, it's generally held that one-third the cost of a new facility is in the building, with the other two-thirds going to the equipment contained inside. Figure the specialized construction means the building itself is 2-4x the cost of basic industrial and you can estimate the cap-ex budget from your local real estate listings. Operational expenditure is harder to get a grasp on but this should give you an idea of the exotic supplies and utility requirements, as well as the sorts of skilled tradespersons you need to staff to keep everything running.
posted by 7segment at 2:44 PM on March 27, 2018

Here’s the microscopy lab I used to run: http://nic.ucsf.edu - if you navigate to the equipment page, each of the microscopes there is between 100 and 500k USD. A high end objective for the microscope can be as much as $15-25k.
Illumina DNA sequencers are in the ballpark of 500-750k. This is why a lot of universities have large core facilities to support all this capital-intensive equipment.
Both of these types of equipment are needed to support cutting edge biomedical research (there might be a publication list on the NIC website as well, if you want to get a more detailed sense of whose doing what).
posted by pombe at 9:09 PM on March 27, 2018

Other people above have covered a lot of the “standard” molecular biology equipment in 2018, but lost in this list of equipment are the things that are... not so standard. For every disease where we’ve identified a particular causative or contributing mutation but we don’t know how it works, it can be useful to make one or more biological models of that disease in a particular cell type or model organism. And each of those living disease models (a mouse, a mutant yeast, a custom cell line) will likely behave slightly differently. Our lab is less than a decade old, and we’re already maintaining freezer collections of 800-some yeast strains and about a thousand different E. coli strains.

Also, maybe this is particularly true in my own work, but is amazing how often the most efficient way to answer a biological question involves doing something that would have been state of the art technology in 1987. Our lab keeps giant sequencing gels around for radio-labeling nucleic acids, because a lot of times we’re working on next generation sequencing methods but sometimes doing something more retro is just the most straightforward approach.

Also our scintillation counter for measuring radioactivity is from 1999, uses user-inserted cards to tell the system which program to select, and is hooked up to a dot matrix printer. I find this delightful.
posted by deludingmyself at 10:12 PM on March 27, 2018 [1 favorite]

I'd echo the idea that lots of labs have stuff that is either produced in extremely small volumes or entirely custom.

I do environmental chemistry, fate, behaviour and effects (on organisms and ecosystems). We have a standard, though well-outfitted lab for this, with around 10-20 million dollars worth of stuff (depending on how you count it). A couple dozen gas chromatographs with various accouterments and attachments, some expensive high-resolution mass spectrometer systems, various physical test equipment, densitometers, rheometers, particle size analyzers, epifluorescent microscopes, automated sample extraction and preparation apparatus, a few shelves worth of small equipment and glass ware. Most of that is off the shelf. That said, we do work with manufacturers to field test some of their newer equipment and instruments for them in a quasi-production setting, providing often a first working example they can point to of their new instruments working on real problems.

But core to much of our work are the simulation rigs and those are all custom. Weathering, the changes chemical undergo in the environment, are key to understanding how these things behave and where they end up and what they can effect. So we have custom temperature control facilities, some we purpose built, some modified from things like fridges, solar exposure systems we've adapted from commercial UV cross-link chambers, hand-built exposure chambers we put out on the roof (one of my most cited papers was done in a glorified chicken coop).

This week, I'm taking delivery of en environmental simulator that can do Arctic conditions to sub-tropical, and from calm sunny day to gale force winds and waves. And while, other groups have systems that are similar, we're pushing into new areas and in being able to simulate more realistically weather regimes---which is a fairly normal progression in any field. It's costing as much as fitting out a new lab would, but hopefully we'll get more than a decade of work out of it. Design of the facility took about three years, and it's taken around 9 months to realize the design and build it. Design included a year-long exercise in identifying the science goals I had, then a couple of years assembling and working with a team of engineers (civil, mechanical, hydraulic, electrical), as well as ice scientists, hydronamicists, physical geographers and limnologists to get the conditions of the tank right. For my part, my team of chemists and physicists working on being able to model and sample the materials in the simulator, and prepare the analytical methods we'd need in order to conduct the experiments we want to do with it. It's all custom, all being built, even the building, to my design. On a not-quite-big science project like this, we've had dozens of people contributing expertise, design, even construction knowledge into the project.

There are many labs out there that work in a service capacity, working up a set of methods which work in a known way to produce a well-defined result. However a lab such as you describe, one that's working to advance the state of the art, generally has to be pushing the technology in some way, and that almost always means custom, new or jury-rigged equipment. It can be a brand new thing you've just got and are trialing with a manufacturer, it can be an older thing you're using in a way or combination that was never intended, or a completely new rig of your own design. But something will be unique to that lab and that research group.
posted by bonehead at 11:06 PM on March 27, 2018

« Older History of Mortified vs Cringe   |   Our housecleaner is going independent. Advice?... Newer »
This thread is closed to new comments.