←
Michael Bohnert3,172 posts
Michael Bohnert avatar

Michael Bohnert

@mbohnert

Defense Researcher / Engineer @RANDCorporation | future tech, drones, air defense, Russia-Ukraine war | Husband, father, skier | Views my own. RT ≠ endorsement.

◷ Joined March 2009▧ 145 Photos and videos
Michael Bohnert avatar
Michael Bohnert@mbohnert·Jul 5

So I got a lot of questions about the A-50U and Russia. Here is an update that explains why these aircraft are so important for air defense. I'll also explain why Russia won't be getting replacements anytime soon Airborne Early Warning and Control (AEW&C) aircraft are basically flying quarterbacks. They are able to see farther than most sensors. They can talk to almost everyone. Their crews can control many assets. This all comes down to understanding the curvature of the Earth. The figure with the Red and Gray "blobs" is an annotated image of the reasonable area a radar can cover. Most radars must have a line of sight to the target. Because of the curvature of the Earth, a surface radar can only see a low flying object 10 to 20 miles away. An airborne radar can potentially see it from over 200 miles away. This has two major implications. The first is it provides much earlier warning. Up to an hour for a strike drone and 10 to 30 minutes for a cruise missile. The second is it covers a much larger area easily replacing a dozen or more surface radars. This means those ground radars can stay turned off until they need to target a missile. A radar that is turned off is much harder to find as there are no radar emissions to detect. This line of sight advantage extends to communications as well. These aircraft have crews specially trained to control air and ground assets. They see and talk to almost everything. Ukraine has spent considerable effort hitting these on the ground and the air. They take a lot of time to replace. The last A-50 was produced around 1992. Even with upgrades over the years, they have been out of production for a while. Note that the base IL-76 airframe is still produced, the radar requires special structures and reinforcements. The replacement, the A-100, has still not entered serial production after over a decade. The radars are very expensive to build and require a lot of specialized parts. In 2025, Ukraine targeted one of the facilities with FP-1 drones, destroying the test aircraft and damaging the factory. Work has likely been on hold since. New aircraft are unlikely for several years. Russia is stuck with the half dozen or so they have for now. Basically just enough to protect Putin and occasionally help Russians fend off increasingly frequent Ukrainian FP-1 and FP-5 Flamingo cruise missiles. The fact Ukraine has sent UK Storm Shadow missiles at near max range into Russia shows how far these A-50U aircraft are kept from the front. Until tomorrow. Have a nice Sunday.

Post attachmentPost attachmentPost attachmentPost attachment

Russia was able to use one of its last remaining A-50U airborne to provide aerial surveillance. Its radar unobstructed by the curvature of the Earth was able to detect Ukrainian FP-5 Flamingo cruise missiles from far away and defeat them. This employment is likely exploitable. Ukraine spent significant effort over the past few years to destroy A-50 aircraft for this very reason. These are officially called "Airborne Early Warning and Control" AEW&C for a reason. There powerful radars can see far and their onboard crew is able to direct aircraft and other assets. While the exact number operating is unknown, it is reportedly in the low single digits. You need 4 to keep one up 24/7 and with 7 you can almost keep 2 up 24/7. Given Russia is now being tracked as using these, Ukraine can plan operations around them where before they had to take the risk the sircraft were in the air. Now that Russia is using them, Ukraine's drone strikes will be less effective due to early warning, but Ukraine can also plan around it. It may be a wash in terms of air defense.

Michael Bohnert avatar
Michael Bohnert@mbohnert·Nov 17

While we keep seeing more companies and organizations have software failures, we keep hearing concerns about cyber or AI attacks. However, most of these appears to be internal code issues. Let's be serious, last year every company was switching to AI coding and reducing coding staff. The models were given more trust than they deserved, many devs still use AI incredibly inefficiently in a vibe-code style, and frankly many of the subject experts were fired, quit, or retired. 12-24 months later we are seeing a string of small internal issues. This is exactly what one would expect from compounding "AI slop". Now that agents can function closer to a junior coder, the world has a collective tech debt. We have already been cleaning up decades of bandaids and "human slop". The next few years will be about rooting out code issues created last year and this year prior to the late summer agents. It's a shame that the past 2 years weren't spent learning to use AI tools and cleaning up the decades of human slop. Thankfully, my team was a little slow to embrace AI. Now, we are using agents to clean up the tech debt and are more productive than ever with consistent quality.

Michael Bohnert avatar
Michael Bohnert@mbohnert·Nov 15

Refinery by refinery. Now that several that are historically strucked haven't been, yet.

Post attachment
Michael Bohnert avatar
Michael Bohnert@mbohnert·Nov 15

Russian Refinery update. Ukrianian strikes have taken a big bite out of Russian refining capacity and ending Russia's brief period of excess supply. Russian production is now into the "gold band" i.e. the region where demand normally sits. We are seeing the first pulls with occasional shortages reported. Given general exhaustion and fuel prices, panic buying may be lessened simply due to a combination of complacency and ability. I want everyone to pay very close attention to the "blue bands". Those are error bands in the method. With open information, only so much specificity is possible. The current Ukrianian strikes are much more impactful with 5-7 hits per strike versus 1-2. As better information becomes available, I expect the average (dark blue line) trends down faster with revised data. Here is the kicker, there are 11 refineries sitting at near max capacity that haven't been hit in several months. Given how much more effective Ukrainian strikes have been, there is a lot of downside to this graph. At current strike progress another .7-1.0 mbpd is possible by next week. Russia is currently sitting at around 3.7 mbpd supply. 3.2 mbpd is about the limit where excess demand has fallen away. Ukrianian strikes have never pushed it below 3.1 mbpd. Regardless, by this time next week Russia may be 10% or more below its demand. I doubt imports can shift fast enough. If I worked at one of those 11 refineries I would find an excuse to call in sick. Full refinery by refinery status in the first reply.

Post attachment
Michael Bohnert avatar
Michael Bohnert@mbohnert·Nov 11

Methodology article.

Michael Bohnert @mbohnert Sep 24 Refinery Output Calculator Methodology Abstract This article describes the methodology behind a daily estimate of how much crude oil is still being distilled at Russian refineries. The estimate is built from a fixed list of forty-one named

Michael Bohnert avatar
Michael Bohnert@mbohnert·Nov 11

Even though Ukraine struck 9 refineries in 11 days, we haven't seen a major change in overall refinery levels. Several Russian refineries that had been spared strikes for several months likely came back online over the past 2 weeks. However, there aren't any left in that category. This is likely why we are hearing reports of localized outages. The system has to adjust to a rapid shift of supply as ~500 kbpd came online and ~600 kbpd when offline in rather concentrated areas. Going forward, expect a more direct downtrend with every strike. However, don't expect significant systemic shortages until the solid line drops into the gold band or outages until it reaches the bottom of the band. Note that a 9 hard strikes per 11 days is likely significantly more than Russia's repair capacity. Restarts take an average off 45-60 days to recover from major hits. Any rate above 0.5 refineries per day have a very noticeable impact. Based on September-October outage maps such as those from @MalcontentmentT , ~5 more strikes in a week would probably make it very colorful. Refinery list is in the first reply and methodology in the second. And as a reminder, this is based on an experimental process being tested real-time. It uses only public sources that can be delayed or noisy such as a restart/shutdown only being announced 2 weeks late. Further, correlating sources such as NASA FIRMS satellite data has some seasonal and environmental quirks. Expect variations in the record when I post updates every ~2 weeks.

Post attachment
Michael Bohnert avatar
Michael Bohnert@mbohnert·Nov 10

I've been working on new updates to my oil tracker. Here is a preview. The version 3 integrated satellite tracking and I showcased it last week. Version 4 will add direct damage assessments from footage. I'm developing a workflow that takes OSINT footage, correlates it to 3D models of drones and locations, works back the trajectories, and helps me identify where drones actually hit and what they struck. This will allow me to assess and estimate what was hit even if the actual hit was obscured. This is from the June 16/17/18 strikes on the Moscow Oil Refinery. Original video in the replies Still a work in progress, but hoping to add soon. As I am testing, feel free to post videos/links that provide the date, location, and a video of a strike that you want to see a 3D recreation of. It helps prioritize beta testing.

0:14
Michael Bohnert avatar
Michael Bohnert@mbohnert·Nov 10

I wanted to make a video to show the effects of drone speed on air defenses. I used the Oct 30 Novokuibyshevsk refinery strikes as my example. I have three drones approach the refinery and pass a tower with a Pantsir. The first is an FP-1 at cruising speed. The second is an FP-5 at Banderol speed for comparison. The third is an FP-5 at cruising speed. You see the approach from the view of the Pantsir. You see the pass from below. The final scene is the FP drone flying away towards the refinery stacks in the distance from the Pantsir's perspective. As you can see, speed make significant difference in the leadup. One other note is how hard it is to see the drones with their profile. I want to be clear. The drones, the Pantsir, the tower, and the refinery stacks are all to scale. I just wanted to give everyone a representation of the speed. Apologies, I didn't quite get the engine sound right for the FP-5 doppler shift. I am starting to recreate all refinery strikes from social media footage to support battle damage assessment to update by oil tracker. Feel free to post a link to a video source for any strike you would like recreated. An example is in the replies. Screenshot and map in the replies.

0:48
Michael Bohnert avatar
Michael Bohnert@mbohnert·Nov 9

The Russian Army is heavily recruiting in Jamaica and the Caribbean. Yesterday a friend that lives in Jamaica send this. A pushed Facebook ad in his feed. I would have reached out to the Jamaican Gov to help save their citizens but I couldn't find an active social media page. The Russian Army uses non-ethnic Russians as disposable troops, drone fodder. Anyone rucruited will be lucky to last a minth. @ChrisO_wiki maybe your next deep dive? @jakluge a price point update @finkd is this a ToS violation for ads? Apologies for the non-standard post and mention spamming, but this will be an actual life or death situation for many.

Post attachment
Michael Bohnert avatar
Michael Bohnert@mbohnert·Nov 4

This is the future of area air defense against low speed threats. A medium/long endurance drone can defend a much larger airspace than any fixed or minimally mobile ground system

Bayraktar rolls out their FPV mothership

0:33
Michael Bohnert avatar
Michael Bohnert@mbohnert·Nov 4

Using cruise missiles to shoot down cruise missiles is an incredibly inefficient use of a military asset. Ukrainska Pravda, quoting Serhii Sternenko: electric interceptors should reach 550-600 km/h for a few minutes. DroneXL puts jet Geran-4/5 at 500-600 km/h. That is an improvement but leaves little or no margin in a tail chase. Early warning and positioning will continue to be extra critical. The cost efficient solution are small turbojet intercept drones. The constraint is turbojet engines and cueing, not airframes. For several weeks, Ukrainain media has discussed using Bars cruise missiles as interceptors for jet drones. They lack the maneuverability to be efficient and their range and payload is literal overkill. I'm really surprised how behind the curve Ukraine is on countering jet drones. European supporters should be ringing alarm bells as there is still no answer to a threat Russia could easily turn on them. While China does manufacture the majority of small jet engines, after more than a year production of small jets should have started elsewhere. A fair amount of my direct research on drones has been spent with hobbyists at the local RC airfield. Ukraine's other innovations have been considered surprisingly quick.

Post attachment