A publication by Onur Oncer // EST. 2026
The Signal Report.
Signal, not noise. Science and world events, sourced and explained: defense tech and electronic warfare, lab science, supplement reality checks, luxury home security, and energy storage. Every claim cited. Where I have a stake, the report says so, and what I link is what my research and experience point to. How this works →
Reports 181 · page 1 of 19
What an RNA integrity number measures
The RIN on your sequencing report feels like a measurement. It is a small neural network trained in 2006 to imitate how experts graded 1,208 ribosomal RNA traces. A good tool, but in a controlled test, blood samples still averaging RIN 7.9 had already shifted 4% of their genes. How to use the number without letting it be the judge.
Read the report → Report 180 · Energy Storage · 2026-10-06 · 7 minWhy a burned battery can catch fire again
On 18 September, about 1,200 still-charged modules at Moss Landing caught fire on their own, twenty months after the 2025 blaze. A fire doesn't drain a battery. The NTSB found responders can't measure or remove that stranded energy from a wrecked pack, and salt-water discharge corrodes terminals and lets voltage rebound.
Read the report → Report 179 · Luxury Home Security · 2026-10-05 · 8 minWhat a perimeter alarm rate leaves out
Estate perimeter systems are sold on detection and false-alarm numbers. A federal prison-perimeter spec shows what those numbers hide: false and nuisance alarms are different things, a sensor alarming ten times an hour drops out of the alarm rate into downtime, and detection has to be proven on site. What to ask before you buy.
Read the report → Report 178 · Defense Tech · 2026-10-05 · 8 minWhat a GPS-jamming map actually shows
A red hex on a public GPS-jamming map is not a jammer. It is airplanes grading their own navigation over a day, at altitudes where a ground jammer can be hundreds of kilometers away. An electronic-warfare officer on what the maps measure, what they filter out, and the aircraft that drop off them entirely.
Read the report → Report 177 · Supplements · 2026-10-04 · 7 minWhat a probiotic's CFU number counts
"50 billion CFU" is a count of colonies that grew on agar plates in one lab. FDA requires probiotics to be listed by weight, which includes dead cells; the CFU figure rests on 2018 draft guidance. Lab-to-lab limits run up to 0.5 log, about a factor of three, and a 2021 review found more than 40% of products that declared a count fell short. Four questions for any CFU claim.
Read the report → Report 176 · AI in the Lab · 2026-10-04 · 7 minWhat 37,000 AI agents actually proposed
A Stanford "virtual biotech" of AI agents, published in Science, proposed a B7-H3 antibody-drug conjugate for lung cancer, and the coverage said the swarm found a drug. A B7-H3 ADC has been in human trials since November 2019. The 37,000 agents were mostly reading trial records. What the AI really added is a fibroblast hypothesis nobody has tested at a bench.
Read the report → Report 175 · Lab Science · 2026-10-03 · 5 minWhat "99% pure by HPLC" means
The purity line on a certificate of analysis usually reports the main peak as a share of what a UV detector could see. Water, salts and silica make no peak, and impurities are weighted by how strongly they absorb. One catechin standard read 100.0% by HPLC and 95.8% by quantitative NMR.
Read the report → Report 174 · Energy Storage · 2026-10-03 · 6 minWhat a 20,000-cycle battery claim means
Hithium's launch release for its grid sodium-ion cell measured 94.2% capacity after 4,000 cycles and said it was expected to pass 20,000 at 70% health. A straight line through that one point lands right on 20,000. Nine months later the claim read simply 'over 20,000 charge cycles.' How to read any five-digit cycle count.
Read the report → Report 173 · Luxury Home Security · 2026-10-02 · 7 minWhat a lock's grade actually measures
Grade 1 is best. On BHMA's residential label, A is best. On high-security cylinders, C is best. The biggest difference between commercial grades is how many cycles the lock survives, and picking and drilling live in a separate standard. What each rating tests, and what none of them cover.
Read the report → Report 172 · Defense Tech · 2026-10-02 · 7 minWhat a bomb robot demo doesn't prove
One clean run on video shows a bomb-squad robot can do a task once. The NIST-DHS-ASTM test methods ask for 10 straight successes, or 20 with one fault, or 30 with three, driven from out of sight. Three clean runs from a robot that is only 80% reliable still gives you a coin flip.
Read the report →