Zlata Karpushina is a go-to-market strategist at FJ Dynamics. On 2DIAMR, Zlata Karpushina took apart the “one-button SLAM scanner” promise and talked about expectation management, real-world use cases and robots.
Episode
1 h 48 minPodcast
The one-button SLAM scanner myth – real-world use cases with FJ Dynamics
Zlata Karpushina, Mehdi Boroumand · FJ Dynamics
In Zlata’s words
“It is one of the most dangerous promises in technology because it affects one of the most important parts of marketing: expectation management.”
“The best scanner is the one that your team can actually use in your real projects.”
Quoted in our tech guides
Will AI and robots replace 3D scanning operators and surveyors? What 9 practitioners say →
- “I think that's what gradually might be happening.”▶ 101:20
Gaussian splatting vs photogrammetry: what 10 practitioners say →
- “Visualization alone only solves half of the problem.”▶ 95:02
Is a 3D scan a digital twin? What 9 practitioners say →
- “For them, 3D scanning creates value by turning the physical building into a reliable digital record.”▶ 10:43
Do you need a licensed surveyor for 3D scanning? What 8 practitioners say →
- “So I think it's the right tool for many survey-grade tasks, and I don't think that the SLAM scanner should compete with terrestrial laser scanners here.”▶ 38:35
SLAM vs terrestrial laser scanning: what 8 practitioners say →
- “Its value is speed, mobility, and flexibility.”▶ 38:48
- “If I need millimeter-level precision for a very controlled survey, I will probably choose a terrestrial scanner.”▶ 39:05
- “It is one of the most dangerous promises in technology because it affects one of the most important parts of marketing: expectation management.”▶ 1:36
- “I feel like in several years companies will be battling on software, not on hardware.”▶ 25:27
Questions Zlata answered
- What causes drift and misalignment in SLAM laser scanning?
- Zlata Karpushina of FJ Dynamics says SLAM performance depends on the environment, capture route, operator, sensors and processing; long corridors, repetitive spaces, glass, reflective surfaces, open areas and poor capture methods can still cause drift and misalignment. ▶ 4:26
- Should a beginner buy a SLAM scanner to scan a large industrial site?
- Per Zlata Karpushina, check with the dealer or manufacturer support that your site conditions suit that specific scanner, because SLAM algorithms have different limitations, and plan for training – no 3D scanner works well without it. ▶ 18:54
- How should you compare SLAM scanners before buying one?
- Zlata Karpushina suggests first defining the accuracy your workflow really needs, then comparing how smooth each scanner’s software workflow is, and finally comparing the whole ecosystem rather than data sheets alone. ▶ 26:43
- How does 3D Gaussian splatting fit into a SLAM scanning workflow?
- Zlata Karpushina sees 3DGS as a visualization layer on top of laser and SLAM data that makes spaces easier to understand for non-technical stakeholders; FJ Dynamics develops it, but as one step of the workflow, not a stand-alone feature. ▶ 92:34
Talks about
Related guests
- Mehdi Boroumand – SLAM scanning practitioner, Australia
- Michael Schoelkopf – Director, Product Management Reality Capture, NavVis
- Ben Williams – COO, Exyn Technologies
- Dylan Gorman – Drone mapping, PilotByte Labs
- Rami Tamimi – Geomatics project manager and Ph.D. candidate, WSP
- Barry Bassnett – Digital twin practitioner, author of Tower of Twins